- Email: [email protected]
Annual serum PICP and ICTP and antler growth in female reindeer (Rangifer tarandus tarandus)
Comparative Biochemistry and Physiology Part B 122 (1999) 111 – 117
Annual serum PICP and ICTP and antler growth in female reindeer (Rangifer tarandus tarandus) Heidi Vierimaa a, Mirja-Liisa Sassi b, Eija Eloranta a, Markku Rahiala c, Jouni Timisja¨rvi a, Seppo Saarela d, Juha Risteli b,* a Department of Physiology, Uni6ersity of Oulu, FIN-90220 Oulu, Finland Department of Clinical Chemistry, Uni6ersity of Oulu, FIN-90220 Oulu, Finland c Department of Mathematical Sciences/Statistics, Uni6ersity of Oulu, FIN-90570, Oulu, Finland d Department of Biology, Uni6ersity of Oulu, FIN-90571 Oulu, Finland b
Received 3 July 1998; received in revised form 27 October 1998; accepted 23 November 1998
Abstract Annual cycle of type I collagen formation and degradation and antler growth was studied in six adult female reindeer, Rangifer tarandus tarandus. Blood samples were collected twice a week during 1 year. Antler length was measured weekly during the antler growth period. An assay for human PICP, the carboxyterminal propeptide of type I procollagen, was used as an indicator of type I collagen formation and an assay for bovine ICTP, the carboxyterminal telopeptide of type I collagen, as an indicator of type I collagen degradation. PICP was clearly increased during the antler growth period. Also ICTP was slightly elevated during antler growth, but the highest values were found in autumn and winter. Our statistical analysis revealed that changes in lagged values (from 3 to 6 weeks) of PICP could be linked to the subsequent changes in the growth rate of the antler, although the highest values of PICP were found during the final third of antler growth. ICTP had significant predictive power as well, but the connection with the growth rate seemed more immediate than that of PICP. In conclusion, antler collagen synthesis can be predicted by PICP, but also ICTP was related to the antler growth. © 1999 Elsevier Science Inc. All rights reserved. Keywords: Antler; Bone growth; Collagen; ICTP; PICP; Reindeer; Remodeling; Linear regression model; Osteoporosis
1. Introduction The growth of antlers in family Cervidae represents the most rapid bone formation in the animal kingdom. For example in male reindeer (Rangifer tarandus tarandus), antlers grow even several centimeters per week during the period of the rapid growth. The antlers of Cervidae are also the only mammalian appendages to be lost and completely regenerated every year [11]. Female reindeer cast their hard antlers normally at the end of April. If the animals have been pregnant, they cast them after parturition in June. After this new antlers begin to grow. Growth is at first slow, but is * Corresponding author. Tel. + 358-40-5909512; fax: + 358-8344522; e-mail: [email protected].
followed by a period of rapid growth which then slows down. Thus, the growth curve is sigmoidal in shape [16]. Antler formation stops at the end of August, when velvet shedding begins. As a whole, the growth period takes about 100 days. Considerable attention has been focused on the endocrine involvement of the antler growth cycle, but only few studies have been carried out on biochemical indicators of bone or collagen metabolism. The previously used markers are alkaline phosphatase, hydroxyproline or osteocalcin [1–4,9,21,23]. The antlers are mainly composed of type I collagen [17,19]. The aim of our study was to measure type I collagen formation and degradation via its specific biochemical markers in relation to the antler cycle. In this study we used the carboxyterminal propeptide of type I
0305-0491/99/$ - see front matter © 1999 Elsevier Science Inc. All rights reserved. PII: S 0 3 0 5 - 0 4 9 1 ( 9 8 ) 1 0 1 5 3 - 0
112
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
Fig. 1. The antler length and serum PICP during 1 year in every individual, which are indicated by numbers 1 – 6. The antler length is indicated by a sole line and PICP by black dots.
procollagen (PICP) as a marker of type I collagen formation and the cross-linked carboxyterminal telopeptide of type I collagen (ICTP) as a marker of type I collagen degradation. In many situations with increased turnover the serum PICP concentration is closely related to the rate of histomorphometrically assessed bone matrix formation and the ICTP to the rate of bone resorption [10]. In addition, both the PICP and ICTP have been shown to closely reflect childhood growth [7]. The data were analyzed by means of a linear variance component model in order to study the dynamic connections between the PICP and ICTP markers and the growth rate of the antler.
2. Materials and methods
2.1. Animals Six adult (3–5 years old) female reindeer (Rangifer tarandus tarandus) were studied during 1996 and 1997.
They were housed outdoors in the Zoological Gardens, Department of Biology, University of Oulu. All the animals were fed ad libitum with a pelleted reindeer feed mix (Poron-Herkku, Raisio, Finland) containing 0.9% Ca, 0.6% P, 2000 U vitamin D3, 10 000 U vitamin A, 10.5% protein, 4.1% fat and 14.5% fiber and other essential trace elements in winter. In summer the food was the same except containing only 8000 U vitamin A. In addition, during the summer the animals were fed with leaves of trees (Salix sp., Betula sp.). Throughout the year they also got lichens and had free access to drinking water. The length of the antler was measured once a week from the pedicle to the tip calculating the tines together. The first measurement was done on the 20th of May, when every antler had started to grow and was long enough to be measured. The last measurement was done on the 20th of September, when the growth had stopped and velvet shedding had started. Blood samples were collected twice a week by jugular venepuncture using evacuated glass tubes (Venoject, Terumo Corporation, Leuven, Belgium). Blood samples
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
113
Fig. 2. The antler length and serum ICTP during 1 year in every individual, which are indicated by numbers 1 – 6. The antler length is indicated by a sole line and ICTP by black dots.
were centrifuged for 15 min at 2000 × g and the serum was stored at −20°C until assayed.
2.2. Blood assays of PICP and ICTP The assay for the human PICP developed by Melkko et al. [15] was applied in this study after the serial dilution curve revealed parallelism between the reindeer serum and the human standard antigen. The PICP was Table 1 Antler growth and serum PICP and ICTP during two periods (mean 9S.E.) Antler growth period Growth in length 1.1 9 0.2 (cm day−1) (n = 108, S.D. = 1.7) PICP (mg l−1) 297.19 16.7 (n = 114, S.D. =178.0) ICTP (mg l−1) 17.4 90.8 (n =114, S.D. = 9.0)
Winter period
122.499.0 (n= 102, S.D. = 91.0) 16.89 0.9 (n=102, S.D. =8.8)
measured using commercially available reagents (Orion Diagnostica, FIN-90460 Oulunsalo, Finland). 100 ml aliquots of standards or serum samples were incubated with 200 ml of the iodinated tracer solution (about 50 000 counts min − 1) and 200 ml of diluted antiserum (in 0.5% normal rabbit serum) for 2 h at 37°C. Then 500 ml of the solid-phase second antibody suspension (20 ml of goat anti-rabbit immunoglobulin antiserum and 150 g of PEG (MW 6000) in 1 l of PBS containing 0.04% Tween 20) was added to each tube and vortexmixed. After 30 min at + 4°C the bound fraction was separated by centrifugation (2000× g, 30 min, +4°C). The supernatant containing the unbound tracer was decanted and the radioactivity of the precipitate containing the bound tracer was counted (Clinigamma 1272, LKB Wallac, Turku, Finland). The inter- and intra-assay variations for PICP are 5 and 3%, respectively. Because the ICTP assay [18] based on human antigen did not show enough cross-reactivity with the reindeer samples, the corresponding bovine ICTP assay estab-
114
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
Fig. 3. p-Values of the PICP connection to the growth as a function of lag in weeks in the linear regression model. Different lags in ICTP shown as separate curves.
lished previously [13,20] was used. The reindeer serum was found to give nearly but not exactly parallel inhibition curve with the bovine standard antigen (data not shown). The procedure was similar to that of the PICP determination, except that the sample volume must be 200 ml possibly due to difference between bovine and reindeer serum antigens. The intra-assay variation of the assay is 5.6%, and inter assay variation about 10%.
2.3. Statistical analysis The antler cycle was divided into two periods: the antler growing period (from 20th of May to 20th of September) and the winter period (from 21st of September to 19th of May). Because of the small sample size, the differences between the mean values of PICP and ICTP of the two periods could not be tested. All the values of every individual were included to describe the data in the figures. This is recommended with a small sample size to minimize the loss of information [14]. The interrelationships between the PICP and ICTP and the antler growth rate (Dlength/week)were analyzed within the following type of models: Dlength =IND + aLPICP +bLICTP + E, week where IND denotes a random intercept allowing for individually varying growth rates and E a measurement specific error term. Both of these components were
taken as random variables, whereas a and b were taken as fixed parameters. For further information concerning the analysis of longitudinal data, see for instance [8]. The symbols LPICP and LICTP denote different lagged values of the concentrations of the markers in the blood samples. This means that 0–8 weeks earlier data of PICP or ICTP was put to the model to predict the present growth. Only two explanatory variables at a time were included in the model because of the sparsity of the data. Anyway, this kind of simple model ought to be able to reveal the nature of the dynamic effects and the possible delay times of the explanatory variables. The parameters were estimated by the maximum likelihood method by means of the MIXED procedure in the SAS statistical software system.
3. Results
3.1. Antler growth Cumulative increase in antler length in every individual together with PICP and ICTP data is shown in the Figs. 1 and 2. As seen in Fig. 1, there were remarkable differences in growth velocity and final antler length between the six individuals. The mean length (9S.E.) of final hard antlers was 184 9 30 cm (n= 6, S.D. =73). The length growth pattern tended to follow sigmoidal growth curve varying by seasons as shown in Fig. 1.
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
115
Fig. 4. p-Values of the ICTP connection to the growth as a function of lag in weeks in the linear regression model. Different lags in PICP shown as separate curves.
Growth in length averaged 1.4 cm day − 1 (range 0.5– 2.3 cm day − 1) when calculating the tines together. The main shaft grew in average about 0.3 cm day − 1.
3.2. Serum PICP The year around serum PICP values in each animal are presented in Fig. 1. They were clearly elevated during the summer peaking from late July to mid-September, when the antler growth rate started to attenuate. Soon after this the values decreased to basic level. During the winter the concentration of serum PICP was the lowest (Table 1). No effect of the abscission of antlers during April is seen on PICP (Fig. 1). The over year mean (9 S.E.) of all reindeer serum PICP values were 225912 mg l − 1 (human reference values 38–202 mg l − 1 for men and 50 – 170 mg l − 1 for women).
3.3. Serum ICTP The year around serum ICTP levels increased gradually during the summer and were at the highest level during fall after the antlers were shed (Fig. 2). Lower values were found in late winter and early spring (Table 1, Fig. 2). The effect of the abscission of antlers on serum ICTP seems unclear. The over year mean value ( 9 S.E.) of serum ICTP was 17.3490.60 mg l − 1.
3.4. Linear regression model In the framework of the mixed model described above, no connection could be found between PICP and the simultaneous growth rate. As shown in Fig. 3, the lagged values of PICP (from 3 to 6 weeks), however, did have significant predictive power. ICTP, on the other hand, seemed to have a fairly immediate connection to the growth rate, but lagged effects were also found. All the p-values connected to the ICTP at different lags were below 0.05 regardless of the lag of the PICP (Fig. 4). The best fitting combinations of the delayed PICP and ICTP values are indicated by bold face text in Table 2. The best combinations seem to be the ones where both PICP and ICTP are delayed several weeks. The causal relationships between the explanatory variables and the growth rate were assumed to be one-sided. As far as PICP is concerned, this assumption seems fairly obvious. On the other hand, taking possible feedback relation between ICTP and the growth rate into account would have complicated the analysis tremendously. To the best of our knowledge, proper statistical methodology for studying feedback relation in the context of longitudinal data with individual effects does not exist yet. In the framework of the one of the best fitting models with L4ICTP and L5PICP as explanatory variables, the
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
116
Table 2 −2 Log Likelihood values of different combinations when delaying serum PICP and ICTP values in the regression model (delay in weeks indicated by the number after PICP or ICTP)
ICTP0 ICTP1 ICTP2 ICTP3 ICTP4 ICTP5 ICTP6 ICTP7 ICTP8 a
PICP0
PICP1
PICP2
PICP3
PICP4
PICP5
PICP6
PICP7
PICP8
207.23 209.29 202.72 208.53 201.33 201.74 202.70 195.00a 202.91
207.32 209.93 202.45 209.55 202.38 202.26 203.25 195.65a 203.21
205.81 208.13 201.34 207.66 202.07 201.34 201.72 194.57a 202.42
204.16 206.78 198.82a 206.16 199.16a 201.22 200.26 193.14a 201.47
206.37 208.96 201.43 207.88 200.92 202.46 203.79 195.34a 202.98
205.39 206.79 199.41a 206.34 192.69a 199.8a 202.59 194.65a 201.68
200.11 204.14 195.06a 203.01 195.83a 196.28a 198.14a 192.73a 199.69a
207.71 209.92 203.15 208.87 202.26 202.54 203.51 195.99a 203.86
206.15 208.77 201.32 208.50 200.74 201.28 202.50 194.74a 202.87
The best combinations (smaller value).
S.D. of the individual random effect OND was estimated as 0.7372. The estimates of the fixed parameters were a = 0.002976 and b =0.09449 corresponding to t-values 2.49 and 4.34.
4. Discussion Type I collagen is the main component of both the mature and developing antler tissue [17,19]. For that reason a rapidly growing antler offers a good opportunity to study changes in type I collagen metabolic markers at serum level. In this study the serum PICP increased remarkably during the antler growth period as expected (Fig. 1). In addition, we observed that changes in the serum PICP are reflected as increased growth about three weeks later (Fig. 3). Several previous studies concerning childhood growth conclude that these markers are able to predict increasing growth much earlier than can be observed by straight length measurements. For this reason they are of great value for example following patients with growth disorders [22]. However, the maximum values of PICP are seen during the final third of antler growth. Similar results during antler growth have been obtained by serum alkaline phosphatase [9,5,23], although the maximum values occurred slightly earlier than in PICP. The increased alkaline phosphate levels have been suggested to indicate increased osteoblastic activity in the growing antler. Another marker, serum osteocalcin indicates antler mineralization. Van Der Eems et al. [23] and Baksi and Newbrey [1] measured an increase in blood osteocalcin during the antler growth and mineralization period. Also hydroxyproline levels have been shown to increase during this time [23], which is due to increased degradation of collagens. Among our animals there was one clear exception (number 2). It had the biggest antlers, but serum PICP elevation was the smallest. In some cases, for example in thyreotoxicosis, when bone turnover is highly in-
creased, serum PICP values are low possibly because of increased elimination of PICP [6]. This may also explain why the PICP values have a maximum after the most rapid antler growth. Also part of the rise in PICP may be derived from the turnover of the skeleton. It is highly probable that the cyclic osteoporosis has some effect on serum PICP during the antler growth as well as on ICTP. ICTP, type I collagen degradation marker, closely reflected growth (Fig. 4) as indicated by previous studies (see e.g. [7]). In humans ICTP is derived from the increased collagen degradation especially during rapid bone turnover. ICTP can also come from skeleton, if calcium is depleted in serum during antler mineralization phase [10]. Increased ICTP levels after velvet shedding period may be due to this. Several research groups have documented a cyclical osteoporosis in skeletal bones in deer related to antler mineralization. They have hypothesized that calcium needed for mineralization of antlers is taken from the skeletal bones and this calcium is later replenished by food sources [4,12]. However, cyclic osteoporosis in the ribs observed by Hillman et al [12] seemed to occur at the time of maximum antler growth and thus it does not necessarily explain the increased ICTP values in late fall and winter. Baksi and Newbrey found plasma levels of parathyroid hormone and 25-hydroxyvitamin D to be significantly higher in summer and to increase progressively during the rapid antler growth period [1,2]. They also observed bone loss in the ribs during antler growth. This also supports increased demand for calcium in antler mineralization phase. In contrast to this, Chao et al. measured calcitonin in white-tailed deer (Odocoileus 6irginianus) and suggested that increased calcitonin levels during rapid antler growth may have prevented bone resorption [5]. In conclusion, our results show that both the PICP and bovine ICTP are able to predict future bone growth in rapidly developing antler, but there are still other factors affecting the serum pool of these markers.
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
Acknowledgements We gratefully acknowledge the expert technical assistance of Ms Helka Koisti and Mr Jari Ylo¨nen for helping with the animals. We are also indebted to Mr Risto Bloigu for valuable discussions concerning the statistics. This study was supported in part by the Technology Development Centre, Finland and Oulu University Hospital.
References [1] Baksi SN, Newbrey JW. Plasma calcemic hormones in mature female reindeer, Rangifer tarandus. Gen Comp Endocrinol 1988;69:262 – 6. [2] Baksi SN, Newbrey JW. Bone metabolism during antler growth in female reindeer. Calcif Tissue Int 1989;45:314–7. [3] Bjarghov RS, Fjellheim P, Hove K, Jacobsen E, Skjenneberg S, Try K. Nutritional effects on serum enzymes and other blood constituents in reindeer calves (Rangifer tarandus tarandus). Comp Biochem Physiol 1976;55A:187–93. [4] Brown RD, Cowan RL, Griel LC. Correlation between antler and long bone relative bone mass and circulating androgens in white-tailed deer (Odocoileus 6irginianus). Am J Vet Res 1978;39:1053 – 6. [5] Chao CC, Brown RD, Deftos LJ. seasonal levels of serum parathyroid hormone, calcitonin and alkaline phosphatase in relation to antler cycles in white-tailed deer. Acta Endocrinol 1984;106:234 – 40. [6] Charles P, Hasling C, Risteli L, Risteli J, Mosekilde L, Eriksen EF. Assessment of bone formation by biochemical markers in metabolic bone disease: separation between osteoblastic activity at the cell and tissue level. Calcif Tissue Int 1992;51:406– 11. [7] Crofton PM, Wade JC, Taylor MRH, Holland CV. Serum concentrations of carboxyl-terminal propeptide of type I procollagen, amino-terminal propeptide of type III procollagen, crosslinked carboxyl-terminal telopeptide of type I collagen, and their interrelationships in schoolchildren. Clin Chem 1997;43:1577 – 81. [8] Diggle P, Liang K-Y, Zeger S. Analysis of Longitudinal Data. Oxford: Oxford University Press, 1994. [9] Eiben B, St. Scharla, Fischer K, Schmidt-Gayk H. Seasonal variations of serum 1,25-dihydroxyvitamin D3 and alkaline phosphatase in relation to the antler formation in the fallow deer (Dama dama L.). Acta Endocrinol 1984;107:141–4.
.
117
[10] Eriksen EF, Charles P, Melsen F, Mosekilde L, Risteli L, Risteli J. Serum markers of type I collagen formation and degradation in metabolic bone disease: Correlation with bone histomorphometry. J Bone Miner Res 1993;8:127 – 32. [11] Goss RJ. Deer antlers. Regeneration, Function, and Evolution. New York: Academic Press, 1983:316. [12] Hillman JR, Davis RW, Abdelbaki YZ. Cyclic bone remodeling in deer. Calc Tissue Res 1973;12:323 – 30. [13] Liesegang A, Sassi, M-L, Risteli, J, Eicher, R, Wanner, M, Riond, J-L. Comparison of different bone resorption markers during hypocalcemia in dairy cows. J Dairy Sci 1998, in press. [14] Matthews JNS, Altman DG, Campbell MJ, Royston P. Analysis of serial measurements in medical research. Br Med J 1990;300:230 – 5. [15] Melkko J, Niemi S, Risteli L, Risteli J. Radioimmunoassay of the carboxyterminal propeptide of human type I procollagen. Clin Chem 1990;36:1328 – 32. [16] Muir PD, Sykes AR, Barrel GK. Growth and mineralisation of antlers in red deer (Cer6us elaphus). New Zealand J Agric Res 1987;30:305 – 15. [17] Price JS, Oyajobi BO, Nalin AM, Frazer A, Russell RGG, Sandell LJ. Chondrogenesis in the regenerating antler tip in the red deer: Expression of collagen types I, IIA, IIB, and X demonstrated by in situ nucleic acid hybridization and immunocytochemistry. Dev Dyn 1996;205:332 – 47. [18] Risteli J, Elomaa I, Niemi S, Novamo A, Risteli L. Radioimmunoassay for the pyridinoline cross-linked carboxy-terminal telopeptide of type I collagen: a new serum marker of bone collagen degradation. Clin Chem 1993;39:635 – 40. [19] Rucklidge GJ, Milne G, Bos KJ, Farquharson C, Robins SP. Deer antler does not represent a typical endochondral growth system: immunoidentification of collagen type X but little collagen type II in growing antler tissue. Comp Biochem Physiol B Biochem Mol Biol 1997;118:303 – 8. [20] Risteli J, Sassi M-L, Eriksen H, Niemi S, Mansell J, Risteli L. Immunochemical characterization of the epitope of the assay for the human carboxyterminal telopeptide of type I collagen (ICTP), Abstract. J Bone Min Res 1997;12(Suppl 1):583. [21] Sempere AJ, Bubenik GA, Smith JH. Antler cycle and thermolabile and thermostabile alkaline phosphatase in white-tailed deer; circannual and circadian rhythms and variation after thyroxine, dexamethasone and ACTH administration. Acta Endocrinol 1986;111:133 – 9. [22] Tapanainen P, Knip M, Risteli L, Kemppainen L, Ka¨a¨r M-L, Risteli J. Collagen metabolites in the prediction of response to GH therapy in short children. Eur J Endocrinol 1997;137:621–5. [23] Van Der Eems KL, Brown RD, Gundberg CM. Circulating levels of 1,25 dihydroxyvitamin D, alkaline phosphatese, hydroxyproline, and osteocalcin associated with antler growth in whitetailed deer. Acta Endocrinol 1988;118:407 – 14.
Annual serum PICP and ICTP and antler growth in female reindeer (Rangifer tarandus tarandus) Heidi Vierimaa a, Mirja-Liisa Sassi b, Eija Eloranta a, Markku Rahiala c, Jouni Timisja¨rvi a, Seppo Saarela d, Juha Risteli b,* a Department of Physiology, Uni6ersity of Oulu, FIN-90220 Oulu, Finland Department of Clinical Chemistry, Uni6ersity of Oulu, FIN-90220 Oulu, Finland c Department of Mathematical Sciences/Statistics, Uni6ersity of Oulu, FIN-90570, Oulu, Finland d Department of Biology, Uni6ersity of Oulu, FIN-90571 Oulu, Finland b
Received 3 July 1998; received in revised form 27 October 1998; accepted 23 November 1998
Abstract Annual cycle of type I collagen formation and degradation and antler growth was studied in six adult female reindeer, Rangifer tarandus tarandus. Blood samples were collected twice a week during 1 year. Antler length was measured weekly during the antler growth period. An assay for human PICP, the carboxyterminal propeptide of type I procollagen, was used as an indicator of type I collagen formation and an assay for bovine ICTP, the carboxyterminal telopeptide of type I collagen, as an indicator of type I collagen degradation. PICP was clearly increased during the antler growth period. Also ICTP was slightly elevated during antler growth, but the highest values were found in autumn and winter. Our statistical analysis revealed that changes in lagged values (from 3 to 6 weeks) of PICP could be linked to the subsequent changes in the growth rate of the antler, although the highest values of PICP were found during the final third of antler growth. ICTP had significant predictive power as well, but the connection with the growth rate seemed more immediate than that of PICP. In conclusion, antler collagen synthesis can be predicted by PICP, but also ICTP was related to the antler growth. © 1999 Elsevier Science Inc. All rights reserved. Keywords: Antler; Bone growth; Collagen; ICTP; PICP; Reindeer; Remodeling; Linear regression model; Osteoporosis
1. Introduction The growth of antlers in family Cervidae represents the most rapid bone formation in the animal kingdom. For example in male reindeer (Rangifer tarandus tarandus), antlers grow even several centimeters per week during the period of the rapid growth. The antlers of Cervidae are also the only mammalian appendages to be lost and completely regenerated every year [11]. Female reindeer cast their hard antlers normally at the end of April. If the animals have been pregnant, they cast them after parturition in June. After this new antlers begin to grow. Growth is at first slow, but is * Corresponding author. Tel. + 358-40-5909512; fax: + 358-8344522; e-mail: [email protected].
followed by a period of rapid growth which then slows down. Thus, the growth curve is sigmoidal in shape [16]. Antler formation stops at the end of August, when velvet shedding begins. As a whole, the growth period takes about 100 days. Considerable attention has been focused on the endocrine involvement of the antler growth cycle, but only few studies have been carried out on biochemical indicators of bone or collagen metabolism. The previously used markers are alkaline phosphatase, hydroxyproline or osteocalcin [1–4,9,21,23]. The antlers are mainly composed of type I collagen [17,19]. The aim of our study was to measure type I collagen formation and degradation via its specific biochemical markers in relation to the antler cycle. In this study we used the carboxyterminal propeptide of type I
0305-0491/99/$ - see front matter © 1999 Elsevier Science Inc. All rights reserved. PII: S 0 3 0 5 - 0 4 9 1 ( 9 8 ) 1 0 1 5 3 - 0
112
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
Fig. 1. The antler length and serum PICP during 1 year in every individual, which are indicated by numbers 1 – 6. The antler length is indicated by a sole line and PICP by black dots.
procollagen (PICP) as a marker of type I collagen formation and the cross-linked carboxyterminal telopeptide of type I collagen (ICTP) as a marker of type I collagen degradation. In many situations with increased turnover the serum PICP concentration is closely related to the rate of histomorphometrically assessed bone matrix formation and the ICTP to the rate of bone resorption [10]. In addition, both the PICP and ICTP have been shown to closely reflect childhood growth [7]. The data were analyzed by means of a linear variance component model in order to study the dynamic connections between the PICP and ICTP markers and the growth rate of the antler.
2. Materials and methods
2.1. Animals Six adult (3–5 years old) female reindeer (Rangifer tarandus tarandus) were studied during 1996 and 1997.
They were housed outdoors in the Zoological Gardens, Department of Biology, University of Oulu. All the animals were fed ad libitum with a pelleted reindeer feed mix (Poron-Herkku, Raisio, Finland) containing 0.9% Ca, 0.6% P, 2000 U vitamin D3, 10 000 U vitamin A, 10.5% protein, 4.1% fat and 14.5% fiber and other essential trace elements in winter. In summer the food was the same except containing only 8000 U vitamin A. In addition, during the summer the animals were fed with leaves of trees (Salix sp., Betula sp.). Throughout the year they also got lichens and had free access to drinking water. The length of the antler was measured once a week from the pedicle to the tip calculating the tines together. The first measurement was done on the 20th of May, when every antler had started to grow and was long enough to be measured. The last measurement was done on the 20th of September, when the growth had stopped and velvet shedding had started. Blood samples were collected twice a week by jugular venepuncture using evacuated glass tubes (Venoject, Terumo Corporation, Leuven, Belgium). Blood samples
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
113
Fig. 2. The antler length and serum ICTP during 1 year in every individual, which are indicated by numbers 1 – 6. The antler length is indicated by a sole line and ICTP by black dots.
were centrifuged for 15 min at 2000 × g and the serum was stored at −20°C until assayed.
2.2. Blood assays of PICP and ICTP The assay for the human PICP developed by Melkko et al. [15] was applied in this study after the serial dilution curve revealed parallelism between the reindeer serum and the human standard antigen. The PICP was Table 1 Antler growth and serum PICP and ICTP during two periods (mean 9S.E.) Antler growth period Growth in length 1.1 9 0.2 (cm day−1) (n = 108, S.D. = 1.7) PICP (mg l−1) 297.19 16.7 (n = 114, S.D. =178.0) ICTP (mg l−1) 17.4 90.8 (n =114, S.D. = 9.0)
Winter period
122.499.0 (n= 102, S.D. = 91.0) 16.89 0.9 (n=102, S.D. =8.8)
measured using commercially available reagents (Orion Diagnostica, FIN-90460 Oulunsalo, Finland). 100 ml aliquots of standards or serum samples were incubated with 200 ml of the iodinated tracer solution (about 50 000 counts min − 1) and 200 ml of diluted antiserum (in 0.5% normal rabbit serum) for 2 h at 37°C. Then 500 ml of the solid-phase second antibody suspension (20 ml of goat anti-rabbit immunoglobulin antiserum and 150 g of PEG (MW 6000) in 1 l of PBS containing 0.04% Tween 20) was added to each tube and vortexmixed. After 30 min at + 4°C the bound fraction was separated by centrifugation (2000× g, 30 min, +4°C). The supernatant containing the unbound tracer was decanted and the radioactivity of the precipitate containing the bound tracer was counted (Clinigamma 1272, LKB Wallac, Turku, Finland). The inter- and intra-assay variations for PICP are 5 and 3%, respectively. Because the ICTP assay [18] based on human antigen did not show enough cross-reactivity with the reindeer samples, the corresponding bovine ICTP assay estab-
114
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
Fig. 3. p-Values of the PICP connection to the growth as a function of lag in weeks in the linear regression model. Different lags in ICTP shown as separate curves.
lished previously [13,20] was used. The reindeer serum was found to give nearly but not exactly parallel inhibition curve with the bovine standard antigen (data not shown). The procedure was similar to that of the PICP determination, except that the sample volume must be 200 ml possibly due to difference between bovine and reindeer serum antigens. The intra-assay variation of the assay is 5.6%, and inter assay variation about 10%.
2.3. Statistical analysis The antler cycle was divided into two periods: the antler growing period (from 20th of May to 20th of September) and the winter period (from 21st of September to 19th of May). Because of the small sample size, the differences between the mean values of PICP and ICTP of the two periods could not be tested. All the values of every individual were included to describe the data in the figures. This is recommended with a small sample size to minimize the loss of information [14]. The interrelationships between the PICP and ICTP and the antler growth rate (Dlength/week)were analyzed within the following type of models: Dlength =IND + aLPICP +bLICTP + E, week where IND denotes a random intercept allowing for individually varying growth rates and E a measurement specific error term. Both of these components were
taken as random variables, whereas a and b were taken as fixed parameters. For further information concerning the analysis of longitudinal data, see for instance [8]. The symbols LPICP and LICTP denote different lagged values of the concentrations of the markers in the blood samples. This means that 0–8 weeks earlier data of PICP or ICTP was put to the model to predict the present growth. Only two explanatory variables at a time were included in the model because of the sparsity of the data. Anyway, this kind of simple model ought to be able to reveal the nature of the dynamic effects and the possible delay times of the explanatory variables. The parameters were estimated by the maximum likelihood method by means of the MIXED procedure in the SAS statistical software system.
3. Results
3.1. Antler growth Cumulative increase in antler length in every individual together with PICP and ICTP data is shown in the Figs. 1 and 2. As seen in Fig. 1, there were remarkable differences in growth velocity and final antler length between the six individuals. The mean length (9S.E.) of final hard antlers was 184 9 30 cm (n= 6, S.D. =73). The length growth pattern tended to follow sigmoidal growth curve varying by seasons as shown in Fig. 1.
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
115
Fig. 4. p-Values of the ICTP connection to the growth as a function of lag in weeks in the linear regression model. Different lags in PICP shown as separate curves.
Growth in length averaged 1.4 cm day − 1 (range 0.5– 2.3 cm day − 1) when calculating the tines together. The main shaft grew in average about 0.3 cm day − 1.
3.2. Serum PICP The year around serum PICP values in each animal are presented in Fig. 1. They were clearly elevated during the summer peaking from late July to mid-September, when the antler growth rate started to attenuate. Soon after this the values decreased to basic level. During the winter the concentration of serum PICP was the lowest (Table 1). No effect of the abscission of antlers during April is seen on PICP (Fig. 1). The over year mean (9 S.E.) of all reindeer serum PICP values were 225912 mg l − 1 (human reference values 38–202 mg l − 1 for men and 50 – 170 mg l − 1 for women).
3.3. Serum ICTP The year around serum ICTP levels increased gradually during the summer and were at the highest level during fall after the antlers were shed (Fig. 2). Lower values were found in late winter and early spring (Table 1, Fig. 2). The effect of the abscission of antlers on serum ICTP seems unclear. The over year mean value ( 9 S.E.) of serum ICTP was 17.3490.60 mg l − 1.
3.4. Linear regression model In the framework of the mixed model described above, no connection could be found between PICP and the simultaneous growth rate. As shown in Fig. 3, the lagged values of PICP (from 3 to 6 weeks), however, did have significant predictive power. ICTP, on the other hand, seemed to have a fairly immediate connection to the growth rate, but lagged effects were also found. All the p-values connected to the ICTP at different lags were below 0.05 regardless of the lag of the PICP (Fig. 4). The best fitting combinations of the delayed PICP and ICTP values are indicated by bold face text in Table 2. The best combinations seem to be the ones where both PICP and ICTP are delayed several weeks. The causal relationships between the explanatory variables and the growth rate were assumed to be one-sided. As far as PICP is concerned, this assumption seems fairly obvious. On the other hand, taking possible feedback relation between ICTP and the growth rate into account would have complicated the analysis tremendously. To the best of our knowledge, proper statistical methodology for studying feedback relation in the context of longitudinal data with individual effects does not exist yet. In the framework of the one of the best fitting models with L4ICTP and L5PICP as explanatory variables, the
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
116
Table 2 −2 Log Likelihood values of different combinations when delaying serum PICP and ICTP values in the regression model (delay in weeks indicated by the number after PICP or ICTP)
ICTP0 ICTP1 ICTP2 ICTP3 ICTP4 ICTP5 ICTP6 ICTP7 ICTP8 a
PICP0
PICP1
PICP2
PICP3
PICP4
PICP5
PICP6
PICP7
PICP8
207.23 209.29 202.72 208.53 201.33 201.74 202.70 195.00a 202.91
207.32 209.93 202.45 209.55 202.38 202.26 203.25 195.65a 203.21
205.81 208.13 201.34 207.66 202.07 201.34 201.72 194.57a 202.42
204.16 206.78 198.82a 206.16 199.16a 201.22 200.26 193.14a 201.47
206.37 208.96 201.43 207.88 200.92 202.46 203.79 195.34a 202.98
205.39 206.79 199.41a 206.34 192.69a 199.8a 202.59 194.65a 201.68
200.11 204.14 195.06a 203.01 195.83a 196.28a 198.14a 192.73a 199.69a
207.71 209.92 203.15 208.87 202.26 202.54 203.51 195.99a 203.86
206.15 208.77 201.32 208.50 200.74 201.28 202.50 194.74a 202.87
The best combinations (smaller value).
S.D. of the individual random effect OND was estimated as 0.7372. The estimates of the fixed parameters were a = 0.002976 and b =0.09449 corresponding to t-values 2.49 and 4.34.
4. Discussion Type I collagen is the main component of both the mature and developing antler tissue [17,19]. For that reason a rapidly growing antler offers a good opportunity to study changes in type I collagen metabolic markers at serum level. In this study the serum PICP increased remarkably during the antler growth period as expected (Fig. 1). In addition, we observed that changes in the serum PICP are reflected as increased growth about three weeks later (Fig. 3). Several previous studies concerning childhood growth conclude that these markers are able to predict increasing growth much earlier than can be observed by straight length measurements. For this reason they are of great value for example following patients with growth disorders [22]. However, the maximum values of PICP are seen during the final third of antler growth. Similar results during antler growth have been obtained by serum alkaline phosphatase [9,5,23], although the maximum values occurred slightly earlier than in PICP. The increased alkaline phosphate levels have been suggested to indicate increased osteoblastic activity in the growing antler. Another marker, serum osteocalcin indicates antler mineralization. Van Der Eems et al. [23] and Baksi and Newbrey [1] measured an increase in blood osteocalcin during the antler growth and mineralization period. Also hydroxyproline levels have been shown to increase during this time [23], which is due to increased degradation of collagens. Among our animals there was one clear exception (number 2). It had the biggest antlers, but serum PICP elevation was the smallest. In some cases, for example in thyreotoxicosis, when bone turnover is highly in-
creased, serum PICP values are low possibly because of increased elimination of PICP [6]. This may also explain why the PICP values have a maximum after the most rapid antler growth. Also part of the rise in PICP may be derived from the turnover of the skeleton. It is highly probable that the cyclic osteoporosis has some effect on serum PICP during the antler growth as well as on ICTP. ICTP, type I collagen degradation marker, closely reflected growth (Fig. 4) as indicated by previous studies (see e.g. [7]). In humans ICTP is derived from the increased collagen degradation especially during rapid bone turnover. ICTP can also come from skeleton, if calcium is depleted in serum during antler mineralization phase [10]. Increased ICTP levels after velvet shedding period may be due to this. Several research groups have documented a cyclical osteoporosis in skeletal bones in deer related to antler mineralization. They have hypothesized that calcium needed for mineralization of antlers is taken from the skeletal bones and this calcium is later replenished by food sources [4,12]. However, cyclic osteoporosis in the ribs observed by Hillman et al [12] seemed to occur at the time of maximum antler growth and thus it does not necessarily explain the increased ICTP values in late fall and winter. Baksi and Newbrey found plasma levels of parathyroid hormone and 25-hydroxyvitamin D to be significantly higher in summer and to increase progressively during the rapid antler growth period [1,2]. They also observed bone loss in the ribs during antler growth. This also supports increased demand for calcium in antler mineralization phase. In contrast to this, Chao et al. measured calcitonin in white-tailed deer (Odocoileus 6irginianus) and suggested that increased calcitonin levels during rapid antler growth may have prevented bone resorption [5]. In conclusion, our results show that both the PICP and bovine ICTP are able to predict future bone growth in rapidly developing antler, but there are still other factors affecting the serum pool of these markers.
H. Vierimaa et al. / Comparati6e Biochemistry and Physiology, Part B 122 (1999) 111–117
Acknowledgements We gratefully acknowledge the expert technical assistance of Ms Helka Koisti and Mr Jari Ylo¨nen for helping with the animals. We are also indebted to Mr Risto Bloigu for valuable discussions concerning the statistics. This study was supported in part by the Technology Development Centre, Finland and Oulu University Hospital.
References [1] Baksi SN, Newbrey JW. Plasma calcemic hormones in mature female reindeer, Rangifer tarandus. Gen Comp Endocrinol 1988;69:262 – 6. [2] Baksi SN, Newbrey JW. Bone metabolism during antler growth in female reindeer. Calcif Tissue Int 1989;45:314–7. [3] Bjarghov RS, Fjellheim P, Hove K, Jacobsen E, Skjenneberg S, Try K. Nutritional effects on serum enzymes and other blood constituents in reindeer calves (Rangifer tarandus tarandus). Comp Biochem Physiol 1976;55A:187–93. [4] Brown RD, Cowan RL, Griel LC. Correlation between antler and long bone relative bone mass and circulating androgens in white-tailed deer (Odocoileus 6irginianus). Am J Vet Res 1978;39:1053 – 6. [5] Chao CC, Brown RD, Deftos LJ. seasonal levels of serum parathyroid hormone, calcitonin and alkaline phosphatase in relation to antler cycles in white-tailed deer. Acta Endocrinol 1984;106:234 – 40. [6] Charles P, Hasling C, Risteli L, Risteli J, Mosekilde L, Eriksen EF. Assessment of bone formation by biochemical markers in metabolic bone disease: separation between osteoblastic activity at the cell and tissue level. Calcif Tissue Int 1992;51:406– 11. [7] Crofton PM, Wade JC, Taylor MRH, Holland CV. Serum concentrations of carboxyl-terminal propeptide of type I procollagen, amino-terminal propeptide of type III procollagen, crosslinked carboxyl-terminal telopeptide of type I collagen, and their interrelationships in schoolchildren. Clin Chem 1997;43:1577 – 81. [8] Diggle P, Liang K-Y, Zeger S. Analysis of Longitudinal Data. Oxford: Oxford University Press, 1994. [9] Eiben B, St. Scharla, Fischer K, Schmidt-Gayk H. Seasonal variations of serum 1,25-dihydroxyvitamin D3 and alkaline phosphatase in relation to the antler formation in the fallow deer (Dama dama L.). Acta Endocrinol 1984;107:141–4.
.
117
[10] Eriksen EF, Charles P, Melsen F, Mosekilde L, Risteli L, Risteli J. Serum markers of type I collagen formation and degradation in metabolic bone disease: Correlation with bone histomorphometry. J Bone Miner Res 1993;8:127 – 32. [11] Goss RJ. Deer antlers. Regeneration, Function, and Evolution. New York: Academic Press, 1983:316. [12] Hillman JR, Davis RW, Abdelbaki YZ. Cyclic bone remodeling in deer. Calc Tissue Res 1973;12:323 – 30. [13] Liesegang A, Sassi, M-L, Risteli, J, Eicher, R, Wanner, M, Riond, J-L. Comparison of different bone resorption markers during hypocalcemia in dairy cows. J Dairy Sci 1998, in press. [14] Matthews JNS, Altman DG, Campbell MJ, Royston P. Analysis of serial measurements in medical research. Br Med J 1990;300:230 – 5. [15] Melkko J, Niemi S, Risteli L, Risteli J. Radioimmunoassay of the carboxyterminal propeptide of human type I procollagen. Clin Chem 1990;36:1328 – 32. [16] Muir PD, Sykes AR, Barrel GK. Growth and mineralisation of antlers in red deer (Cer6us elaphus). New Zealand J Agric Res 1987;30:305 – 15. [17] Price JS, Oyajobi BO, Nalin AM, Frazer A, Russell RGG, Sandell LJ. Chondrogenesis in the regenerating antler tip in the red deer: Expression of collagen types I, IIA, IIB, and X demonstrated by in situ nucleic acid hybridization and immunocytochemistry. Dev Dyn 1996;205:332 – 47. [18] Risteli J, Elomaa I, Niemi S, Novamo A, Risteli L. Radioimmunoassay for the pyridinoline cross-linked carboxy-terminal telopeptide of type I collagen: a new serum marker of bone collagen degradation. Clin Chem 1993;39:635 – 40. [19] Rucklidge GJ, Milne G, Bos KJ, Farquharson C, Robins SP. Deer antler does not represent a typical endochondral growth system: immunoidentification of collagen type X but little collagen type II in growing antler tissue. Comp Biochem Physiol B Biochem Mol Biol 1997;118:303 – 8. [20] Risteli J, Sassi M-L, Eriksen H, Niemi S, Mansell J, Risteli L. Immunochemical characterization of the epitope of the assay for the human carboxyterminal telopeptide of type I collagen (ICTP), Abstract. J Bone Min Res 1997;12(Suppl 1):583. [21] Sempere AJ, Bubenik GA, Smith JH. Antler cycle and thermolabile and thermostabile alkaline phosphatase in white-tailed deer; circannual and circadian rhythms and variation after thyroxine, dexamethasone and ACTH administration. Acta Endocrinol 1986;111:133 – 9. [22] Tapanainen P, Knip M, Risteli L, Kemppainen L, Ka¨a¨r M-L, Risteli J. Collagen metabolites in the prediction of response to GH therapy in short children. Eur J Endocrinol 1997;137:621–5. [23] Van Der Eems KL, Brown RD, Gundberg CM. Circulating levels of 1,25 dihydroxyvitamin D, alkaline phosphatese, hydroxyproline, and osteocalcin associated with antler growth in whitetailed deer. Acta Endocrinol 1988;118:407 – 14.