Trace Elements in Milk of Guinea Pigs During a 20-Day Lactation1

Trace Elements in Milk of Guinea Pigs During a 2o-Day Lactation 1 RALPH R. ANDERSON Department of Dairy Science University of Missouri Columbia 65211 ABSTRACT

Trace elements have been measured in milk samples at great expense. Most studies have used flame photometry and atomic absorption (17). Automation of atomic absorption has been limited in scope compared with inductively coupled plasma-atomic emission spectroscopy (ICP-AES) (8, 15), a technology enabling the measurement of up to 30 elements at the same time on a single milk sample of 4 mI. The elements considered to be in trace or microconcentrations include those other than the seven major elements in milk: K, Ca, P, Cl, Na, S, and Mg. Changes in these macrominerals, except S, have been reported for guinea pig milk (3). Trace elements of importance in milk include Zo, Cu, Fe, and AI. Heavy metals, such as Ph, Cd, and Hg, have received considerable interest because of their high toxicity and possible contamination in foods such as whole milk and milk products. Although these heavy metal contaminants are detectable by ICP-AES, none was detectable in the present investigation. The purposes of this study were to quantify trace elements in guinea pig milk: and to establish patterns of change, if any, through an entire lactation in the guinea pig.

This study was designed to detennine patterns of change in milk trace elements over a complete lactation. Guinea pigs were selected because of ease of milking and short lactation duration. Common short-hair albino guinea pigs were milked using a milking apparatus for 4 to 6 h every other day following isolation from their offspring. Forty-eight samples of 4 mI each were analyzed for trace minerals using inductively coupled plasma spectroscopy. Observations represented 14 d scattered over 20 d of lactation. Eight elements with sufficiently high quantities to measure in milk were zinc (4.18 ppm), strontium (1.12), aluminum (.81), boron (.79), iron (.71), copper (.56), barium (.23), and manganese (.019). Regression analyses were used on rectilinear, quadratic, exponential and cubic models. Quadratic functions fit the data best in six elements, whereas in manganese the best model was the rectilinear. Indices of detennination (R2) relating to day of lactation varied from R2 = .18 in boron to a high of .62 in strontium. For six elements, concentrations increased as milk volume decreased. The exception was zinc, which decreased in the latter half of lactation. (Key words: trace elements, milk, guinea pig)

MATERIALS AND METHODS

INTRODUCTION

Received November 27, 1989. Accepted April 19, 1990. lContribution from the Missouri Agricultural Experiment Station, 10urnal Series Number 10,961. Approved by the Director. This investigation was supported in part by the University of Missouri Institutional Biomedical Research Grant RR07053 from the National Institutes of Health. 1990 1 Dairy Sci

73:2327-2332

Common albino English short-hair guinea pigs (Camm Research Institute, Wayne, NJ) served as the foundation stock. Guinea pigs were fed pellets (Purina, St. Louis, MO) and were provided with fresh water at all times. Samples of milk were obtained from lactating guinea pigs by means of a milk:ing apparatus designed in our laboratory (13). Samples were placed in acid-washed glass vials and frozen until trace elements analyses were made. Fortyeight milk: samples of 4 mI each representing eight guinea pigs were analyzed for trace elements by ICP-AES at the University of Missouri Trace Substances Center. Milk: samples and standards were digested with 4% perchloric

2327

2328

ANDERSON

acid to serve as the matrix. Two standards were zero concentration and 1000 times the minimum detection limit of the element. Days re~ resented and numbers of samples were: d I, 1; d 2, 5; d 4,4; d 5, 1; d 6, 4; d 8, 5; d 10,4; d 11, 1; d 12,4; d 14,5; d 16,4; d 17, 1; d 18,4; and d 20, 5. Concentrations of samples were reported as parts per million. Conversions were made to micromolar Wlits. Based upon the previously reported nonnal lactation curve in the guinea pig (3), total amounts of trace minerals secreted into milk were calculated. Data on each of eight trace elements (AI, Ba, B, Cu, Fe, MD, Sr, and Zn) were subjected to ANOVA, systems regression, and correlation (14). Components in the ANOVA included the regression model tested and the error term with individual animal effects being part of the error tenn. Regression equations included linear, quadratic, and cubic models. Those in which the highest order tenn was significant were selected to describe changes in milk trace elements during lactation. Scatter points were graphed along with the mean and 95% confidence limits of the regression equation selected. RESULTS

Trace elements measured by ICP-AES included the following 31 elements: Ag, AI, As, B, Ba, Be, Bi, Ca, Cd, Co, Cr, Cu, Fe, K, Li, Mg, MD, Mo, Na, Ni, P, Ph, Sb, Se, Si, Sn, Sr, Ti, TI, V, and Zn. Macrominerals, including Ca, K, Mg, Na, and P, were detected in concentrations and patterns of change equal to those reported previously (4). Eight microminerals or trace elements found at concentrations considerably higher than the minimal detectable limits of the methodology included AI, B, Ba, Cu, Fe, Mn, Sr, and Zn. On a few occasions, trace elements were detected at concentrations slightly higher than the minimal limits of the instrument: Cd at .02 to .m ppm, Cr at .08 ppm, Li at .02 to .06 ppm, Mo at .03 to .05 ppm, Ni at .10 ppm, Ph at .10 to .30 ppm, Se at .40 ppm, Si at .35 to 1.30 ppm, and Ti at .01 ppm. Nondetectable elements and the minimal detectable limits of the instrument included nine trace elements: At at <.02 ppm, As at <.20 ppm, Be at <.002 ppm, Bi at <.30 ppm, Co at <.03 ppm, Sb at <.06 ppm, Sn at <.20 ppm, TI at <.80 ppm, and V at <.02 ppm. Journal of Dairy Science Vol. 73,

No.9, 1990

Sufficient numbers of points throughout the lactation curve enabled us to plot data points and to evaluate statistically by analysis of variance and regression the patterns of change in eight trace elements: AI, B, Ba, Cu, Fe, Mn, Sn, and Zn. Ot the microminerals in guinea pig milk, Zn was in highest concentration at an average of 4.18 ppm or 63.9 ~. The peak concentration was on d 2 at 5.34 pm or 81.6 ~, whereas the lowest concentration was on d 20 at 2.61 ppm or 39.9 ~. Regression analysis revealed the best fit of the data points was a quadratic equation as follows: Y (concentration of Zn in ppm) = 4.92 + .076X - .0093X2 (X = day, n = 48, R2 = .56, P<.OOOI) (Figure 1). Strontium was second in concentration with a mean of 1.12 ppm or 12.8 ~. The lowest concentrations were on d 1 and 2 at .82 ppm or 9.4 ~. Concentrations increased gradually throughout lactation until a high of 1.66 ppm or 18.9 ~ was reached on d 20. The best fit was the quadratic equation: Y (concentration of Sr in ppm) = .916 - .0201X + .0028X2 (X = day, n = 48, R2 = .62, P<.OOOI) (Figure 1). AIuminum was the third highest trace element in concentration with a mean of .81 ppm or 30.0 ~. This element varied greatly in concentration from day to day with a low of .20 ppm or 7.4 ~ on d 5, to 1.18 ppm or 43.7 ~ on d 6, to a high of 1.50 ppm or 55.6 ~ on d 20. Because of the great variations within and among days, the regression equations were not significant. In consideration of its simplicity, the linear model chosen was: Y (concentration of AI in ppm) = .567 + .022X (X = day, n = 48, R2 = .08, P = .33) (Figure 1). The trace element B had the next highest concentration in guinea pig milk. The lowest concentration was on d 10 of lactation at .48 ppm or 44.4 ~, whereas the highest was on d 18 at 1.32 ppm or 122.2 J.IM. Individual daily values varied from .30 to 2.00 ppm. The best fit over the 2O-d lactation period was a quadratic equation: Y (concentration of B in ppm) = 1.017 - .079X + .004X2 (X = day, n = 48, R2 = .11, P<.05) (Figure 1). Iron was next in concentration at .71 ppm or 12.7 ~. Individual values varied from a low of .30 ppm or 5.4 J.IM on d 8 to a high of 1.40 ppm or 25.1 ~ on d 18. The best fit of the data was the quadratic equation: Y (concentration of Fe in ppm) = .746 - .053X + .0035X2

2329

LACTATION CHANGES IN MILK. TRACE ELEMENTS

8

122.0

l

6

91.5

r

I

1.8

. -.-.... •

0.

61.0 ::l'::l.

4

-:;

1.2

....

Vi

C N

2

30.5

a

.6

E

13.6

4

8

12

16

4

8

12

16

20

DAYS Of LACTATION

148

2.0

l -

185

3

111

1.6

l

148

2

74 ::l' ::l.

~

.

:

a 0

.•....

·t.·· . ..

4

,

8

.........

I • :

12

•

::l.

a

a

20

4

4'

::l'

I

6.8

DAYS Of LACTATION

0. 0.

--

.

a 0

27.2

20.4

E

E 0. 0.

2.4

. ............ .

E

0. 0.

1.2

.8

CD

37

.

....•

""

..............

.4

.~._.,~

a 16

20

. . .. . :

~ ..•......... .•....

.

•.... 4········:····· :

111 74 37

O+--...---,----,--,---...,-L-O

o

4

8

12

16

20

DAYS Of LACTATION

DAYS Of LACTATION

Figure 1. Individual data points, regression lines of best fit, and 95% confidence limits for the trace elements Zn. Sr, AI, and Ba in guinea pig milk over a 21M1lactation. Concentrations are graphed as parts per million on the left axis and as micromolarity on the right axis. Equations representing the regression lines are presented in Table I.

(X = day, n = 48, R2 = .49, P<.OOOl) (Figure 2). Based upon this equation, the lowest concentration was .55 ppm on d 8, whereas the highest was 1.09 ppm on d 20. Copper was lower than Fe in concentration at .56 ppm or 8.8 J-lM. Individual values varied from a low of .17 ppm or 2.7 J-lM on d 12 to a high of 1.7 ppm or 26.8 J-lM on d 1. Means tended to be low in the middle of lactation and high at the beginning and end of lactation. The best equation to express the changes was the quadratic function: Y (concentration of Cu in ppm) = 1.042 - .125X + .0057X2 (X = day, n = 48, R2 =.48, P<.OOOl) (Figure 2). According to this equation, the concentration of Cu began on d 1 at .92 ppm and dropped to a low point on d 10 at .34 ppm. It then rose gradually to a high of .82 ppm on d 20. Concentrations of Ba ranged in individuals from a low of .035 ppm or .25 J.lM to a high of .44 ppm or 3.2 J-lM. The mean of 48 observa-

tions was .23 ppm or 1.67 J-lM. Values tended to be high early and late in lactation and low during the middle, similar to Fe and Cu. The best equation for the data points of Ba was: Y (concentration of Ba in ppm) = .335 - .032X + .0016XZ (X = day, n = 48, R2 = .18, P<.Ol) (Figure 2). The trace element in lowest concentration to be measured according to a defmable pattern was Mo. It has a mean of .019 ppm or .35 J.IM for 48 samples. The low individual value was .007 ppm, which was just above the minimal detectable amount of .006 ppm. The high value was .048 ppm. Daily mean values tended to increase from .01 ppm in early lactation to .03 ppm at the end. This was confirmed by the regression analysis, which showed the rectilinear model to be more accurate than the more complex models. It was as follows: Y (concentration of Mn in ppm) = .0115 + .00073X (X = day, n 48, R2 .22, P<.OOl). With this

=

=

Journal of Dairy Science Vol. 73,

No.9, 1990

2330

ANDERSON

2.0

35.0

2.0

31.5

1.5

28.8

1.5

25.2

E 1.2 Q.

18.9

1.2 E Q.

.,

.8 0

"-

.

21.6

• • 0

Q.

~

:>

14.4

.

u

12

I 0

0 8

4

:;.

.8 .4

7.2

0 0

~

~

:;.

,.,

,

• ".

.4

., ···

··· ......

Q.

15

.• .-,

.4

0 0

E

~"""......

.2

....

Q.

~ 0 CD

.1

./~/

~~./

.

..

..

15

12

20

3.50

.048

2.88

.038

,870

.

•...•

.3

8

4

DAYS OF LACTATION

· "'. I

6.3

0

20

DAYS OF LACTATION

.5

12.6

2.15 ~

:;.

1.44

E Q.

.029

. .......::::

~

E

.019

::;

,72

0 +----,----r---,,---.-----,-1-0 4 8 12 15 20 0

...,....

.

,435 ~

.

.•...........•

.218

~.. -

,010

,553

0+----.----,---,,---.-----,-1-0

o

4

8

12

15

20

DAYS OF LACTA TID

DAYS OF LAC ATIO

Figure 2. Individual data points, regression lines of best fit, and 95% confidence limits for the trace elements Fe, Cu, Ba, and Mn in guinea pig milk over a 2Q-d lactation. Concentrations are graphed as parts per million on the left axis and as micromolarity on the right axis. Equations representiDg the regression lines are presented in Table 1.

predictive equation, the low concentration was .012 ppm on d 1 and the high was .026 ppm on d 20 (Figure 2). Predictive equations based upon molarity as well as those for parts per million are presented in Table 1 for each of the eight trace elements. Along with these are the probabilities and the coefficients of determination. Note the superiority of fit to the predictive equations by Zn, Sr, Fe, and Co compared with the other four trace elements. Total amounts of each trace element secreted over the 20-<1 lactation were obtained by multiplying daily concentration, derived from the predictive equation generated, times daily milk production, which was calculated from the predictive equations in a series of three generated previously (3). Based upon these computations, total Zn secreted was 2124 IJ.g, Sr 454 IJ.g, Al 344 IJ.g, B 322 !J.8, Fe 279 IJ.g, Cu 226 IJ.g, Da 94.6 IJ.g, and Mn 8.03 IJ.g. Daily peaks were Journal of Dairy Science Vol. 73,

No.9, 1990

203 IJ.g on d 6 for Zo, 36.7 IJ.g on d 7 for Sr, 29.0 IJ.g on d 7 for AI, 27.7 IJ.g on d 6 for B, 22.4 IJ.g on d 6 for Fe, 20.6 IJ.g on d 5 for Cu, 8.09 IJ.g on d 6 for Ba, and .67 IJ.g on d 7 for Mo. DISCUSSION

Trace elements in guinea pig milk changed significantly along carefully defined patterns during a 20-<1 lactation. These findings indicate complexities in mechanisms controlling individual trace elements in their journey across the mammary epithelial cell from blood to milk. Because the data suggest a unique pattern of change for each trace element, a separate and distinct carrier mechanism probably controlled each one. Concentrations varied to a remarkable degree with Zn being highest at over 4 ppm, Sr at over 1 ppm, Al at .8 ppm, Bat .8 ppm, Fe at .7

2331

LAcrATION CHANGES IN MILK TRACE ELEMENTS

TABLE 1. ~tions of best fit generated for eight trace elements in guinea pig milk on a daily basis for 2Q-d lactation. Element

Equation

P

R2

zu. zu.

4.92 + .076X - .0093X2 75.2 + 1.16X - .142X2 .916 - .020IX + .0028X2 10.5 - .229X + .032X2 .567 + .022X 21.0 + .82X 1.017 - .0792X + .OO4X2 94.2 - .731X + .037X2 .746 - .053X + .0035X2 13.3 - .948X + .063X2 1.042 - .125X + .0057X2 16.4 - 1.97X + .l1X2 .335 - .032X + .0016X2 2.44 - .233X + .012X2 .0115 + .00073X .209 + .0133X

<.0001

.56

<.0001

.62

>.33

.08

<.05

.11

<.0001

.49

<.0001

.48

<.01

.18

<.001

.22

ppm ~

Sr, ppm Sr, J.IM AI, ppm

Al,J.IM B, ppm B, J.IM Fe, ppm Fe, J.IM Cu,ppm Cu, J.IM Ba, ppm Ba, J.IM Mn, ppm Mn, ~

lX = Day of lactation, n = 48, P is level of probability, and R2 is the coefficient of determination.

ppm, Cu at .6 ppm, Ba at .3 ppm, and Mn at .02 ppm. Although some of these were similar, their respective molar equivalents presented differences sufficiently great to deny a common mechanism of transfer. Zinc was second highest at 63.9 IJM and B highest at 73.1 IJM. Third in molarity was Al at 30.0 IJM. Fourth and fifth were Sr at 12.8 IJM and Fe at 12.7 IJM. Copper was sixth at 8.8 IJM, Ba was seventh at 2.2 IJM, and Mn last at less than .4 IJM. Little commonality was observed in these concentrations. Therefore, the conclusion was drawn that each was controlled by a separate and independent mechanism. Directions of change in concentration of each element were varied. Zinc was maintained at a high concentration until d 8 before it began to decrease. Concentrations continued to fall until the end of lactation. The rate of decline was nearly .2 ppm/d or 4% of the total per day. In comparison, lactose fell .27% in concentration or 5.4% of the total each day beginning d 6 (2), and K fell 1.2 mM/d or 3.7% of the total each day (4). Orotic acid declined at a rate of 1.6 ~g/ml or 5% of the total each day of the lactation (18). Percentage declines in concentrations of 3.7 to 5.4% compared favorably with milk volume decreases of 5.0%/d during the declining phase of lactation in the guinea pig (3). H a highly specific protein carrier controlled Zn concentrations in guinea pig milk, it may in tum have been regulated in its synthesis or transfer by K within the secretory cell. Con-

versely, Zn concentrations may have been related to intracellular enzymes because Zn is known to be a cofactor of several important mammary gland enzymes. Zinc concentration in cow milk averages 3.81 ~g/g or ppm (12), which is similar to the magnitude of Zn in guinea pig milk. Zinc is a required element in the function of more than 100 enzymes (to), including carbonic anhydrase, lactic dehydrogenase, and alkaline phosphatase, all of which have been found in milk (19). The amount of Zn in alkaline phosphatase, which is associated with the fat globule membrane, accounts for only a very small amount of total milk Zn, whereas 85% is associated with casein (10). Zinc in milk may be complexed with protein (5, 16), as well as with calcium phosphate and citrate (6). Decreases in concentration of Zn as lactation progresses have been reported in human milk (7, 11). Strontium increased in a quadratic pattern during lactation. This was almost identical to the change in Ca throughout lactation (4). Therefore, the transfer of Sr into milk followed almost identically that of Ca, suggesting a common mechanism. Strontium was present in a ratio of approximately 1:1500 ratio of Sr to Ca throughout lactation. This ratio is considerably higher than that of 1:2500 ratio of Sr to Ca found in human blood (9, 17). Strontium has been reported to be 170 ~g/L or .17 ppm in cow milk (10, 12), much lower than that found in guinea pig milk. It was reported to be 15 Journal of Dairy Science Vol. 73, No.9, 1990

2332

ANDERSON

times as concentrated in guinea pig milk as in human milk (1). Copper and Fe decreased in concentrations during the middle part of lactation, Co declining much more steeply than Fe. At the end of lactation, Fe concentration exceeded that found in early lactation, whereas Co was lower on d 20 than on d 1. Lack of similarity in patterns of change in Co and Fe suggested separate controls in the form of metaUoprotein carriers or secretory cellular enzymes for these elements. Barium and B mimicked Co and Fe in that concentrations began high, decreased in midlactation, and rose again at the end However, because the concentrations in parts per million or molarity were so different between the two elements, no evidence was available from which to speculate on the presence of a common carrier mechanism. Aluminum and Mn changed concentrations rectilinearly; each rose gradually over the 2D-d lactation. The great differences in concentrations between the two elements indicated no link in mechanism of control. Data presented in this study stimulate speculation as to mechanisms of control at the secretory cell and on the need for trace elements at the given concentrations by the neonates consuming them. REFERENCES 1 Anderson, R. R. 1989. Comparison of major minerals and trace elements in cattle, guinea pig and human milb. J. Dairy Sci. 72(Suppl. 1):312. (Abstr.) 2 Anderson, R R., and D. D. Chavis. 1986. Changes in macroingredients of guinea pig milk through lactation. J. Dairy Sci. 69:2268. 3 Anderson, R. R, M. S. Salah, L. G. Sheffield, and W. N. McKenzie, Jr. 1984. Milk production and lactation curve of guinea pigs. J. Dairy Sci. 67:185. 4 Anderson, R. R, and L. G. Sheffield. 1988. Maaominerals in guinea pig milk during 21 days of lactation. J.

Journal of Dairy Science Vol. 73,

No.9, 1990

Dairy Sci. 71:337. 5 Baumy, J. J., and G. Brule. 1988. Effect of pH and ionic strength on the binding of bivalent cations to ll-casein. Lait 68:400. 6 Brusbmiller, J. G., R W. Ames, F. A. Jacobs, and L. S. Nelson. 1989. Intraluminal chemistry of zinc in milks. BioI. Trace IDem. Res. 19:71. 7 Casey, C. E., M. C. Neville, and K. M. Hambidge. 1989. Studies in human lactation: secretion of zinc, copper and manganese in human milk. Am. J. Clin. Nutr. 49:773. 8 Haas, W. J., and V. A. Fassel. 1980. Inductively coupled plasma atomic emission spectroscopy. Page 167 in Elemental analysis of biological materials: current problems and techniques with special reference to trace elements. Tech. Rep. Ser. 197, Int. Atomic Energy Agency, Vienna, Anst. 9 Harrison, G. E., W H.A. Raymond, and H. C. Tretheway. 1955. The metabolism of strontium inman. Clin. Sci. 14: 681. 10 Jenness, R 1988. Composition of milk. Page 12 in Fundamentals of dairy chemistry. N. P. Wong, ed. Van Nostrand Reinhold, Pub!., New York:, NY. 11 Kana, M. V., A. Kirksey, O. Ga1a1, N. S. Bassily, G. H. Harrison, and N. W. Jerome. 1989. Effect of short-term oral zinc supplementation on the concentration of zinc in milk from American and Egyptian women. Nutr. Res. 9: 471. 12 McBean, L. D., and E. W. Speckmann. 1988. Nutritive value of dairy foods. Page 380 in Fundamentals of dairy chemistry. N. P. Wong, ed. Van Nostrand Reinhold, Publ., New York, NY. 13 McKenzie, W. N., Jr., and R. R. Anderson. 1979. A modified device for collecting milk from guinea pigs. J. D~Sci. 62:1469. 14 SAS User's Guide. 1982. SAS Inst., Inc., Cary, NC. 15 Schramel, P. 1983. Consideration of inductively coupled plasma spectroscopy for trace element analysis in the biomedical and environmental fields. Spectrochim. Acta 38B:I99. 16 Singh, H., A. Flynn, and P. F. Fox. 1989. Binding of zinc to bovine and human milk proteins. J. Dairy Res. 56:235. 17 Van Loon, J. C. 1980. Analytical atomic absorption spectroscopy: selected methods. Academic Press, New York:, NY. 18 Wahab, I. M., and R. R. Anderson. 1988. Orotic acid in guinea pig milk: Changes in concentration during lactation. I. Dairy Sci. 71:2554. 19 Whitney, R M. 1988. Proteins of milk. Page ·106 in Fundamentals of dairy chemistry. N. P. Wong, ed. Van Nostrand Reinhold, Pub!., New York:, NY.