Figure 1. Examples of some of the environmental inputs on bone elongation (red). The functional outputs of the skeleton (gray) are kept in physiological balance with the inputs to maintain homeostasis. Movement, for example, is a functional role of the skeleton. Lack of movement (inactivity), however, can cause mineral loss and reduced elongation rate. Final bone length is the product of a complex interplay of genetic and environmental factors that act on multiple levels during postnatal skeletal growth. Temperature is often underrecognized for its contribution to bone lengthening.

Figure 2. Schematic of a long bone growth plate and principal blood supplies. Image at right shows the proximal tibial growth plate from a mouse that was injected with oxytetracycline (OTC) to label newly formed bone. Orientation matches the schematic on the left. The growth plate appears as a dark band between OTC fluorescence in epiphyseal and metaphyseal bone. The perichondrium and vascular network remain intact. The growth plate is comprised of a heterogeneous collection of chondrocytes located between the ossified epiphysis and metaphysis of immature long bones. Growth plates are conventionally subdivided into morphologically and functionally distinct regions: reserve, proliferative, (proliferative‐hypertrophic transition) and hypertrophic zones. Bone elongation occurs through a series of well‐orchestrated events in which chondrocytes in columns divide, mature, and are replaced with mineral at the chondro‐osseous junction where bone‐forming osteoblasts invade from the metaphyseal vasculature (line drawing modified from Serrat et. al, 2009, reference (289), Copyright 2009 American Physiological Society).

Figure 3. Temperature effects on femur length in mice. Representative femora from mice housed at cold (7°C) and warm (27°C) temperatures from weaning age to adulthood showing the effect of warm ambient temperature on extremity lengthening. The underlying cause of such effects is not immediately obvious since homeotherms maintain tightly regulated internal body temperatures independent of their external environment [original text and image, from Serrat et al., 2008, reference (285) p. 19347, Copyright 2008 National Academy of Sciences, U.S.A.].

Figure 4. Caudal end of pigs demonstrating temperature effects on extremity growth in mammals. Littermates were housed at 35°C (left) and 5°C (right) from weaning to maturity. The animals were the same approximate body mass but the cold‐reared pig had a short, stocky build with shortened tail and limbs when compared to its littermate raised at warm temperature. This illustrates the major impact that temperature can have on the phenotype of a growing animal [reproduced, with permission, from Weaver and Ingram, 1969, reference (354), p. 711].

Figure 5. Growth curves demonstrate that mouse tail elongation rate is impacted by ambient rearing temperature without affecting body mass. Tail length (A) and body mass (B) of male CF‐1 mice were measured at weekly intervals beginning at weaning. Tail length was significantly reduced in the cold within a single week (p < 0.05, one‐way ANOVA) and follows a clear temperature gradient. Body mass remained similar among all three groups (p > 0.10). Means ± 1 S.E.M. shown (sample size in legend). These results demonstrate that the temperature growth response is almost immediate and is limited to the extremities (tissue temperatures are shown in Figure 8).

Figure 6. Scatter plot of postmortem tail lengths and femur lengths from male CF‐1 mice housed at 7, 21, or 27°C beginning at weaning (N = 94). Animals were studied at different endpoints between 4.5 and 11.5 weeks age (285). There is a strong positive relationship between endpoint tail and femur length in the aggregate sample that includes all ages and groups (Pearson's correlation r = 0.88, p < 0.01), suggesting that the temperature effect in the limbs is similar to that in the tail. Within‐group correlations were near or above R = 0.90 for each of the three temperature groups (Pearson's R= 0.87 cold; 0.93 control; 0.93 warm; all significant at p < 0.01). The significant correlations indicate that the temperature impact on tail length, which can be measured noninvasively over the course of an experiment, is a good proxy for the effect on limb length [original data, with permission, from Serrat, 2007, reference (282)].

Figure 7. Diagram illustrating some of the major physiological routes through which temperature can influence bone elongation rate. The solid lines indicate direct temperature effects. The dashed lines are intended to show several indirect ways that temperature can alter bone growth. The double‐headed arrows indicate the interactions between these dynamic systems. Many temperature effects on bone lengthening occur by indirect mechanisms involving the endocrine and vascular systems by changing the amount of hormones and nutrients that reach cartilage growth plates, the sites of bone elongation. Other than cartilage canals present in larger mammals, growth plates do not have a direct penetrating blood supply and solutes are delivered to cartilage by transport from the surrounding vasculature. Temperature‐induced changes in blood flow, diffusion, and tissue permeability can impact how systemic and paracrine growth regulators are able to move into and through the cartilage. Heat and cold exposure can also alter the concentration of circulating hormones that are important for regulating longitudinal growth, such as thyroid hormone, leptin, estrogen and glucocorticoids. Direct temperature effects can occur at the cellular level in growth plates by altering expression of genes that control production of receptors, channels, and enzymes involved in cell proliferation, metabolism, matrix production, mineralization, and vascularization at the chondro‐osseous junction. These cellular processes can also indirectly affect growth by changing the physical properties of cartilage and its surrounding blood vessels in such a way that alters the transport of critical nutrients and growth factors. Since blood is also a major source of heat, changes in blood flow could indirectly modulate tissue temperature and elicit direct cellular reactions. This highlights that it is essential to perform whole animal in vivo studies, because the level of complexity of an intact organism with intact circulation cannot be adequately modeled by in vitro studies alone (original artwork by Tom Pickens and Matt Crutchfield, Graphic Designers, Marshall University School of Medicine).

Figure 8. Average peripheral temperature of the ear (left) and tail (right) of 6.5‐week‐old mice show strong positive correlations with their growth (Pearson's r in figure, both correlations significant at p < 0.001). Temperatures were measured weekly using a noncontact thermometer at the tail base and ear. Male CF‐1 mice were housed continuously at 7, 21, or 27°C beginning at weaning (3.5 weeks). Growth curves are shown in Figure 5. Total tail growth was measured as change in length from start. Ear area was measured at the endpoint. After only 3 weeks in the housing conditions, extremity temperature and extremity growth in these littermate mice showed a remarkable covariation with ambient temperature as demonstrated in the scaled cartoon at the top. Note the gradient response to temperature in that mice housed at 21°C were intermediate in all features. These results suggest that there are direct temperature effects on developing cartilaginous tissues in the heterothermic appendages (original artwork, with permission, by Matt Crutchfield, Graphic Designer, Marshall University School of Medicine).

Figure 9. Histological analysis of proximal tibial growth plates from 4.5‐week‐old mice housed for 1 week at 7, 21, or 27°C (N = 11 per group). (A) Representative sections from cold‐ and warm‐housed mice after 1 week of temperature exposure. Sections are stained with Safranin‐O/fast green (cartilage is red and bone is green). Orientation matches Figure 2. Growth plate morphology was unexpectedly similar at cold and warm temperatures. There were no major appreciable differences in overall size, shape, or organization of the cartilage. (B) Tibial growth plate height was slightly smaller in cold‐housed mice when compared to the two warmer temperature groups, but the only significant pairwise comparison was between the 7 and 21°C groups (p < 0.05 by one‐way ANOVA and Tukey's post‐hoc test). The 21 and 27°C groups were statistically indistinguishable. (C) Total length of the tibia differed significantly in a predictable temperature gradient (all pairwise comparisons p < 0.05 by one‐way ANOVA and Tukey's post‐hoc test). Means ± 1 S.E.M. shown. These results suggest that the dynamic events contributing to the bone elongation differences cannot be adequately captured in a postmortem histology snapshot. See text for discussion.

Figure 10. Temperature effects on growth of transplanted mouse tail vertebrae. (A) Kidney with transplanted tail vertebrae from young mice that had been growing for 3 weeks in the warmer abdominal cavity. (B) Examination of cleared and mounted tails revealed that the transplanted vertebrae were over 20% longer and were better vascularized than those left on the tails of age‐matched controls. The transplant environment in the abdominal cavity is much warmer than that of its usual location on the tail, suggesting that direct temperature‐related changes in cell kinetics could be inducing the growth differences [adapted, with permission, from Noel and Wright, 1972, Ref. (220), p. 635‐636].

Figure 11. Effects of incubation temperature on metatarsal growth in vitro. (A) Representative comparison of metatarsals from the same individual (left‐right antimeres) grown in culture at cold and control temperatures for 4 days. Cold significantly impaired metatarsal elongation. (B) Histological sections of left‐right antimeres from 2‐day cultures stained with BrdU showed increased chondrocyte proliferation at warmer temperature, particularly near the rounded distal end of the epiphysis at the top of the image (BrdU positive cells indicated by dark brown stained nuclei). (C) Length‐match comparison of metatarsals from two different individuals grown at cold and warm temperatures for 4 days (stained with Safranin‐O/fast green so cartilage is red and bone is green). These two individuals had a similar percentage increase in length from baseline at 32C (25% relative increase) and 39C (31% relative increase). Note that although the bones were the same total length after the 4‐day culture, the cartilaginous portion of the warm metatarsal was expanded in width and had more extracellular matrix (note the cellular density in the cold metatarsal). The warm metatarsal had clearly grown more in relative width than it had in length. (D) Ratio of matrix area/cell area measured in 150 μm × 100 μm sample regions quantifies the observation in (C), that the cold metatarsal had less matrix and more cells in a given sample area. Boxes show interquartile range (middle 50%), whiskers denote upper 10th and 90th percentile, and horizontal line denotes the median. Sample size (N) listed in graph. The similarity between the control and warm groups stresses again that there is limited information that can be obtained from a static histological analysis of a postmortem growth plate [A and B adapted, with permission, from Serrat et al., 2008, Ref. (285), Copyright 2008 National Academy of Sciences, U.S.A.; C and D from Serrat, 2007 Ref. (282)].

Figure 12. In vivo imaging of bone microvasculature demonstrates cellular‐level resolution. Multiphoton image of blood vessels in the plexus surrounding the tibial (knee) growth plate of a live, anesthetized 5‐week‐old mouse. Orientation matches Figure 2. Vessels were visualized using a multiphoton microscope at 880nm illumination after an intravenous injection of the small (332 Da) tracer fluorescein. Plasma is fluorescent yellow and blood cells appear as dark shadows within the vessels. The collagen‐rich perichondrium around the growth plate (blue‐gray pseudocolor) was visualized by second‐harmonic generation (SHG), a robust signal from unstained collagen that is unique to multiphoton excitation. SHG allows collagenous structures to be identified without injecting stains or dyes. This research examines temperature effects on the vascular‐cartilage interfaces of elongating bones to determine how solutes move out of the vasculature and into cartilage plates, using different‐sized tracers introduced into the systemic circulation. Real‐time imaging using multiphoton microscopy is an innovative advancement in the study of bone elongation because it enables dynamic visualization and quantification of cellular‐level growth plate transport in an intact animal, in a way not possible using other techniques. Imaging was done by Maria Serrat on a Leica TCS SP5 II Broadband Confocal and Multiphoton Microscope housed in the Molecular and Biological Imaging Center at Marshall University.

Figure 13. Multiphoton images of vessels in the subperichondrial plexus from the same mouse with its limb bathed in cool (23°C) and then changed to warm (37°C) lactated Ringer's saline. The growth plate is oriented as in Figure 2. Images were captured 30 min apart at the same growth plate depth and location, verified in a series of optical sections imaged superficial to deep. Vessels were visualized on a multiphoton microscope at 780 nm illumination after an IV injection of fluorescein. Plasma is fluorescent (white) and blood cells appear as dark shadows within the vessels. Vessels were notably enlarged after switching to the warmer temperature (arrows), indicating the sensitivity of the microvasculature to sensing and responding to temperature change [original text and image, with permission, from Serrat et al., 2009, Ref. (289), Copyright 2009 American Physiological Society].

Figure 14. Time‐lapse multiphoton images illustrating fluorescein (yellow pseudocolor) entering the growth plate in vivo within seconds after intra‐cardiac injection. Orientation matches Figure 2. Oxytetracycline (OTC) label (green pseudocolor) of epiphyseal (top) and metaphyseal (bottom) bone facilitates localization of the growth plate, which appears between the white arrowheads in the far left frame. Boxes depict sample regions for quantifying fluorescence in reserve (r), proliferative (p), transitional (t), and hypertrophic (h) subdivisions of the growth plate, as well as in the vasculature (v) of the metaphyseal bone. While the conventional terminology used in Figure 2 is upheld for simplicity, these sample regions were not strictly confined to the standard growth plate zones and boxes may have included cells from adjacent morphological territories [original text and image, with permission, from Serrat et al., 2009, Ref. (289), Copyright 2009 American Physiological Society].

Figure 15. Timed accumulation curve of fluorescein tracer in the growth plate (all subdivisions combined) standardized to 8‐min vascular concentrations. Means ± 1 S.E.M. are shown for each group at 1‐min intervals spanning 10 min after intra‐cardiac injection of fluorescein. Mean values in the warm were nearly double those measured at cool temperature, clearly demonstrating that warm temperature increases solute delivery into the growth plate in vivo. This dynamic process is relevant for understanding how temperature modulates bone elongation. These differences could not be detected using other methods such as the static histology shown in Figures 9 and 11 [original text and image, with permission, from Serrat et al., 2009, Ref. (289), Copyright 2009 American Physiological Society].

Figure 16. Stacked bar graph illustrating the spatial distribution of fluorescein in subdivisions of the growth plate and vasculature 8 min after intra‐cardiac injection. Regions correspond to boxes shown in Figure 14. Each box represents a percentage of the total fluorescence measured among all five regions so that the total shown is 100%. At warm temperature, fluorescein is distributed evenly among all sample areas (∼20% per region). In the cold, fluorescein is distributed nearly equally among growth plate compartments, with slightly more in the transition (18.6%) and hypertrophic (21%) regions compared to reserve (15.7%) and proliferative (15.8%), but a greater percentage of the tracer (28.9%) remains in the cold vasculature when compared to the warm group. These results suggest that the cool temperature inhibited solute transport out of the vasculature, allowing less fluorescein to enter the growth plate matrix [original text and image, with permission, from Serrat et al., 2009, Ref. (289), Copyright 2009 American Physiological Society].