A model calcifying system of primary osteoblast cell cultures derived from normal embryonic chicken calvaria has been flown aboard the shuttle, Endeavour, during the National Aeronautics and Space Administration (NASA) mission STS-59 (April 9–20, 1994) to characterize unloading and other spaceflight effects on the bone cells. Aliquots of cells (∼7 × 106) grown in Dulbecco's modified Eagle's medium (DMEM) + 10% fetal bovine serum (FBS) were mixed with microcarrier beads, inoculated into cartridge culture units of artificial hollow fiber capillaries, and carried on the shuttle. To promote cell differentiation, cartridge media were supplemented with 12.5 μg/ml ascorbate and 10 mM β-glycerophosphate for varying time periods before and during flight. Four cartridges contained cells from 17-day-old embryos grown for 5 days in the presence of ascorbate prior to launch (defined as flight cells committed to the osteoblastic lineage) and four cartridges supported cells from 14-day-old embryos grown for 10 days with ascorbate before launch (uncommitted flight cells). Eight cartridges prepared in the same manner were maintained under normal gravity throughout the flight (control cells) and four additional identical cartridges under normal gravity were terminated on the day of launch (basal cells). From shuttle launch to landing, all cartridges were contained in closed hardware units maintaining 5% CO2, 37°C, and media delivery at a rate of ∼1.5 ml/6 h. During day 3 and day 5 of flight, duplicate aliquots of conditioned media and accumulated cell products were collected in both the flight and the control hardware units. At the mission end, comparisons among flight, basal, and control samples were made in cell metabolism, gene expression for type I collagen and osteocalcin, and ultrastructure. Both committed and uncommitted flight cells were metabolically active, as measured by glucose uptake and lactate production, at approximately the same statistical levels as control counterparts. Flight cells elaborated a less extensive extracellular matrix, evidenced by a reduced collagen gene expression and collagen protein appearance compared with controls. Osteocalcin was expressed by all cells, a result indicating progressive differentiation of both flight and control osteoblasts, but its message levels also were reduced in flight cells compared with ground samples. This finding suggested that osteoblasts subjected to flight followed a slower progression toward a differentiated function. The summary of data indicates that spaceflight, including microgravity exposure, demonstrably affects bone cells by down-regulating type I collagen and osteocalcin gene expression and thereby inhibiting expression of the osteogenic phenotype notably by committed osteoblasts. The information is important for insight into the response of bone cells to changes of gravity and of force in general.
ADAPTATION BY the vertebrate skeletal system in response to external environmental forces is a well-recognized phenomenon.(1–5) Bone will change its architectural geometry or form, for instance, as a result of its loading or unloading through mechanical effects, which include gravity, buoyancy, and other physical influences on the tissue structure. The adaptive phenomenon in bone is not limited to architectural change alone but may extend to changes in the development, metabolic state, and function of the skeleton as a whole. Examples of such skeletal adaptation include increasing bone mass as a result of exercise regimes,(6–10) loss of bone mass during prolonged immobilization(11, 12) and weightlessness,(13) and remodeling of bone tissue during fracture healing.(14) Under the effects of loading or unloading, then the vertebrate skeleton maintains a form and structure appropriate to the applied force(s) through modulation of its architecture, mass, and composition. The muscles, and possibly other tissues and organs in the body, are affected similarly by loading and also undergo adaptation.(1, 15)
Precisely how these changes in bone and other vertebrate tissues are made is not certain although they are thought to be mediated ultimately at the cellular level of structural hierarchy. To understand more completely the basic biology, biochemistry, and physicochemistry of the adaptive events in bone, a model calcifying system of primary osteoblast cell cultures derived from normal embryonic chicken calvaria has been flown aboard Endeavour during the National Aeronautics and Space Administration (NASA) shuttle mission STS-59 (April 9–20, 1994) at the Kennedy Space Center, FL, to assess spaceflight effects, including microgravity, on the bone cells. These cultures have been characterized extensively under normal gravity (1G) in terms of their reproducible cell growth and development, extracellular organic matrix production, andmineralization.(16–18) Thus, by comparing any changes that can be observed in the same cultured cells maintained in the unloaded state during spaceflight, insight and knowledge may be gained concerning its influence on bone cells and the function of the cells under normal gravitational conditions.
This article reports the first data that partially characterize the influence of spaceflight and weightlessness on the cultured embryonic chicken bone cells, age-selected for those either committed or yet uncommitted to the osteoblastic lineage. These two embryonic cell populations were chosen to try to distinguish between an anticipated spaceflight bone loss attributable to lower cellular activity of mature osteoblasts or a decrease in osteoblast differentiation that yields fewer mature cells. Comparisons among the cells subjected to spaceflight or maintained at 1G show that flight affects osteoblasts by down-regulating type I collagen and osteocalcin gene expression and slowing osteoblast progression toward a differentiated function. These results may be fundamentally important in considerations of bone loss experienced by humans and other vertebrates during spaceflight or in cases of extended fracture healing, long-term bed rest and immobilization, and related conditions in normal gravity. The information here also may have significant implications regarding bone maintenance against the effects of osteoporosis and other bone degenerative pathologies. Abstracts of selected previous results have been published.(19, 20)
MATERIALS AND METHODS
The method of culturing cells for the shuttle experiments was accomplished in two stages, one in the laboratories at the Children's Hospital and the other subsequently at the Kennedy Space Center, FL. During each cell preparation, osteoblasts were obtained by trypsin-collagenase digestion of 14- or 17-day-old embryonic chick calvaria and grown in culture to confluence in minimum essential medium Eagle (MEM, Sigma Chemical Co., St. Louis, MO, U.S.A.), supplemented with 10% fetal bovine serum (FBS, Sigma Chemical Co.). The general technique has been described in detail elsewhere.(16) All incubations were carried out at 37°C and 5% CO2. The cells were isolated with trypsin-EDTA, centrifuged, and resuspended in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% FBS. The cells were counted and aliquots (∼7 × 106 cells in 2 ml) were mixed with 125 mg Cytodex 3 microcarrier beads (8 ml in volume; Pharmacia, Uppsala, Sweden), previously hydrated and sterilized. Ten milliliters of cell-bead mixtures were incubated in 10-cm Petri dishes for 24 h to facilitate cell attachment to the microcarriers. Next, approximately 50 milliliters of cells and beads were combined and transferred to sterile 250-ml Costar T-flasks with vented caps (Costar Corp., Cambridge, MA, U.S.A.), to which 200 ml of 10 mM HEPES-buffered DMEM + 10% FBS were added, for transporting to the Kennedy Space Center. At this site (a sterile, fully equipped laboratory in Hangar L of the Space Center), the cell-bead mixtures were replenished immediately with standard carbonate-buffered media and placed overnight in a 5% CO2 incubator. Excess medium was then removed from the flasks on settling of their cell and bead contents. The remaining medium from each flask was combined in a sterile 50-ml centrifuge tube and next divided into 10 aliquots, each in a separate sterile 50-ml centrifuge tube. Medium was added to bring the total volume in the individual tubes to 10 ml, the appropriate volume for subsequent inoculation of cells and beads into cartridge culture units containing artificial hollow fiber capillaries (Cellco Cellmax Quad units; Spectrum Laboratories, Laguna Hills, CA, U.S.A.) that were to be loaded aboard the Endeavour.
Five Cellco Cellmax Quad artificial capillary culture units were used for the incubation and development of the osteoblasts before shuttle launch. The bioreactor units consisted of four independent flowpaths, each having a 250-ml medium reservoir, silicone tubing to provide a flow circuit and gas exchange, and an artificial capillary cartridge assembly. Each combination of a reservoir bottle, tubing rack, and cartridge common to a single flowpath was numbered separately. Continuous medium flow was maintained by a four-position pump and an electronic flowrate control module. Each cylindrical cartridge contained a bundle of hollow polypropylene fibers (630 μm outer diameter and 330 μm inner diameter) arranged in parallel to fill the cartridge shell. Cartridge ends were connected to the inflow and outflow tubing and two sideports were utilized for introduction of the cell-bead mixture and then sealed. Cell-bead inoculum settled onto the outer surfaces of the capillary fibers (the extracapillary space [ECS]) while media passed through the hollow center of the fibers (the intercapillary space [ICS]). Capillary fibers were semipermeable and accommodated direct transport of nutrients to cells while metabolic wastes were removed from the cells and diluted into the circulating medium. Quad units were placed in sterile, humidified, temperature-controlled incubators in which CO2 and other atmospheric gases diffused through the tubing walls.
Two days before inoculation of cells and beads, the sterile water content of all cartridges was replaced by sterile phosphate-buffered saline in the ICS and sterile 5% coating buffer (1 mg/ml human fibronectin [Collaborative Biomedical Products, Bedford, MA, U.S.A.] in sterile water) in the ECS. After 24 h, these solutions were replaced with DMEM + 10% FBS and the cartridges were connected to respective flowpaths, placed in incubators, and equilibrated an additional 24 h. The following day, media in all cartridges and media bottles on all Quad units were removed. New labeled bottles containing 150 ml of fresh media were connected to the systems, and the individual 10-ml cell-bead mixtures were introduced through the sideports into 10 cartridges. Cell-bead mixtures were allowed to equilibrate for 30 minutes in this configuration, after which each pump was activated at a slow speed for circulation of media in respective quad units. Circulation speed was increased by one step after 24 h and 48 h.
In cartridges containing bone cells obtained from 14-day-old embryos, medium was supplemented with 10 mM β-glycerophosphate (Sigma Chemical Co.) 3 days after inoculation of the cell-bead mixture. Two days later, these cartridges also received 12.5 μg/ml ascorbate (Sigma Chemical Co.). Each subsequent cell feeding (at approximately 2-day intervals) consisted of fresh DMEM + 10% FBS, β-glycerophosphate, and ascorbate. Bone cells isolated from 17-day-old embryos received fresh media supplemented with 10 mM β-glycerophosphate 24 h after inoculation and then received ascorbate 2 days later.
Twenty-four hours after the single addition of ascorbate to the cells obtained from 17-day-old chick embryos, cartridges were removed from respective Quad units by clamping inflow and outflow ports and disconnecting flowpaths. Cartridges were then transferred expeditiously to specially designed Space Tissue Loss flight and ground control modules (STL-A). These modules were self-contained cell culture units, incorporating sterile media and media bags, flowpaths, and sump bags (collection bags for spent media and accumulated cell products) that could be loaded aboard the shuttle in a mid-deck locker. Media and sump bags and flowpaths were numbered to correspond to the respective numbered cartridge to which they were connected. The STL-A modules are described in detail elsewhere.(21) At this point, the cells obtained from 14- and 17-day-old embryos had been growing 19 days and 12 days, respectively, from the time of their original isolation. The 10 cartridges containing cells from 14-day-old embryos were divided between flight and ground control STL-A modules, four randomly selected cartridges in each. The 10 cartridges containing cells isolated from 17-day-old embryos were similarly selected randomly and placed in the same flight or control module as their 14-day-old counterparts. Four cartridges, two with cells from 14-day-old embryos and two with cells from 17-day-old embryos, remained connected to a single incubated Quad unit, utilized if necessary as replacements for cartridges in any of the STL-A modules.
The STL-A modules and incubated Quad unit were maintained with fresh media, β-glycerophosphate, and ascorbate as the cartridges were connected and these systems remained stable throughout the subsequent 3-day prelaunch time period for this mission. At launch, the Quad unit was terminated and its four cartridges (termed “basal” for the experimentation and derived from both 14- and 17-day-old embryos) were processed as described below. Flight and control cartridges continued to function until the shuttle landing when each STL-A module was processed in the same manner as that of the basal cartridges. At the launch, each flight and control module contained four cartridges with cells grown for 5 days in the presence of ascorbate (termed “committed” and derived from 17-day-old embryos) and four cartridges with cells grown for 10 days in the presence of ascorbate (termed “uncommitted” and derived from 14-day-old embryos).
Flight and recovery
From the prelaunch time of onboard delivery to the shuttle to the recovery after the flight and landing, all flight and ground control cartridges were maintained under similar conditions in STL-A modules controlling a 5% CO2 atmosphere, ∼1.5 ml/6 h delivery rate of 90 ml total culture media (DMEM + 10% FBS supplemented with 12.5 μg/ml ascorbate and 10 mM β-glycerophosphate), and 37°C. During day 3 and day 5 of flight, duplicate ∼3.0-ml aliquots of the sump fluid were sampled separately by automated means and stored in vials preloaded with protease inhibitors in both the flight and the ground control STL-A modules. The incubator module temperature and CO2 levels were monitored identically. Although no data were recorded for variations in temperature or gas levels, the hardware box signaled out-of-range values (±5%) that were preset for each. Sump fluid collection times were synchronized between flight and controls.
Shuttle landing for this mission occurred at Dryden Flight Center, Edwards Air Force Base, CA. On retrieval of the STL-A module (∼45 minutes from landing to the beginning of processing), the flight cartridges, their flowpaths, and media and sump bags were disconnected in a laminar flow hood. Photographs (35 mm) were immediately made of these components to record their physical condition, media color, and relative volume. Control cartridges were treated at the Kennedy Space Center at the same time and in the same manner as flight cartridges (events coordinated by speaker phones between the Dryden and Kennedy laboratories). Basal cartridges had earlier been processed identically. For the planned molecular biological and biochemical assays, all numbered media and sump bag fluid contents were measured by uptake into calibrated syringes and then transferred to correspondingly numbered sterile 50-ml centrifuge tubes, which were frozen under liquid nitrogen. Sump fluid aliquots, collected for both flight and control cartridges during day 3 and day 5 of the mission, were frozen in like manner at this time. For four of the flight cartridges (two with committed cells and two with uncommitted cells) and similarly for four of the control cartridges, fluid contents in the respective flow paths and ECS of the Cellco culture units were removed by uptake into calibrated syringes and measured in volume, syringe barrels were capped, and syringes were frozen under liquid nitrogen. A 0.5% glutaraldehyde-2% paraformaldehyde mixture in 0.05 M cacodylic acid, pH 7.4, 4°C, was then quickly injected by syringe into each of these four flight and four control cartridges to fix the contents for subsequent scanning and transmission electron microscopy. For the remaining four flight and four control cartridges, the fluid contents in flowpaths and ECS were again removed with syringes, measured, and frozen. Each of these cartridges, devoid of fluid, was then also frozen under liquid nitrogen for further biochemical analyses. All frozen materials were maintained in a −80°C freezer until they were shipped under dry ice to Children's Hospital. Cartridges treated with glutaraldehyde-paraformaldehyde were processed for electron microscopy as detailed below.
Assays of glucose and lactate levels in fluids recovered from cartridges, flowpaths, cartridge reservoirs, and/or sump bags from committed and uncommitted basal, ground control, and flight specimens were made using a Kodak IBI Biolyzer and Kinetic Module (Eastman Kodak Co., New Haven, CT, U.S.A.). The unit accommodated individual dry slides onto which 10 μl of sample were deposited with a calibrated pipette. Colorimetric sensitivity for glucose and lactate ranged from 0.2 to 4.5 g/liter and from 0.5 to 12.0 mmol/liter, respectively, with the Biolyzer unit. Results were compared statistically using analysis of variance (ANOVA) and a post hoc multiple comparison test by the Bonferroni method and statistical values were represented as ±1 SD of calculated mean values.
Isolation, extraction, and analysis of RNA
Intact cartridges were frozen under liquid nitrogen, shipped from the retrieval site under dry ice, and stored at −80°C until processed. Total RNA and protein contents of cartridges were isolated using Tri-Reagent (Molecular Research Center, Inc., Cincinnati, OH, U.S.A.). Recovery of RNA was compared by a two-sample unpaired Student's t-test. Isolation generally followed the manufacturer's instructions, but several modifications were made to optimize RNA and protein extractions. All tubes, instruments, and other equipment were cleaned and treated to be RNAase-free for the procedures. Each cartridge was opened rapidly with a clean polyvinylchloride (PVC) pipe cutter and its complete frozen contents of capillary fibers and attached bone cells and matrix were transferred to a 50-ml polypropylene centrifuge tube. A 4-ml volume of cold (4°C) Tri-Reagent was added to each tube and fibers were teased apart with a sterile dental probe. Fibers and Tri-Reagent were then vortex-mixed 30–60 s at room temperature. Next, 2 ml more of cold Tri-Reagent were added and mixed (∼20 s). Tubes stood at room temperature (5 minutes) after which they were centrifuged (3000 rpm for 2 minutes). As much Tri-Reagent as possible was recovered with a pipette and transferred to clean polypropylene or autoclaved glass tubes.
Tri-Reagent supernatants were processed further by the addition of 1 ml chloroform/tube. Tubes were vortexed and allowed to stand for 15 minutes at room temperature after which they were centrifuged (10,000 rpm for 20 minutes at 4°C). The homogenate separated into three phases after this step, RNA being in the aqueous layer, DNA in the interphase, and protein in the organic layer. The aqueous fraction of each respective separation was recovered with a pipette and transferred to a sterile tube. Isopropanol (0.5 ml isopropanol/ml Tri-Reagent recovered) was added to each tube and vortex-mixed (30 s at room temperature). The solution was then divided equally into individual Eppendorf tubes (five tubes for this set of experiments), allowed to stand for 20 minutes in an ice-filled container, and centrifuged (12,000 rpm for 10 minutes at 4°C). Supernatant was decanted from the resulting RNA pellet and the pellet was rinsed with cold (4°C) 75% ethanol. Ethanol was decanted and tubes were placed on ice for an additional 30 minutes after which they were centrifuged 10 s and the remaining ethanol was removed. Tubes were replaced on ice (20 minutes) and each pellet was redissolved in ∼10 μl RNAase-free water. Contents of the five Eppendorf tubes were then pooled to yield a sample of ∼50 μl. The concentration of total RNA recovered from the extraction procedure was determined by measuring optical absorption at (OD260).
RNA was resolved on 1% denaturing agarose gels containing 2.2 M formaldehyde.(22, 23) Five micrograms of total RNA were loaded in each gel lane. Loading of RNA was verified by ethidium bromide staining of the gel before blotting onto Zeta Probe membranes (Bio Rad Corp., Richmond, CA, U.S.A.).32P-radiolabeled complementary DNA (cDNA) probes were synthesized(24, 25) and hybridization was carried out as described previously.(26, 27) cDNA for pro-α1(I)(28) and chicken osteocalcin(29) was used to examine levels of expression for messenger RNA (mRNA). Quantitation of autoradiograms was made using an LKB Ultra II scanning densitometer (LKB, Broma, Sweden) and values were normalized to 18S ribosomal RNA (rRNA) obtained by hybridization of the same blot to a probe containing conserved nucleotide sequence of the 18S rRNA (Ambion, Inc., Austin, TX, U.S.A.). Three independent measurements were made for each of the blots of type I collagen and osteocalcin.
A hand-held Hirox Micro Hi-Scope video microscope (Hirox Co., Ltd., Tokyo, Japan) was used to obtain ×20-×100 magnified images of intact cartridges, cartridge contents, or specimens resin-embedded in dishes (see below). The microscope was interfaced with a video monitor connected to a digital image recorder to produce hard copies of sample areas of interest.
Scanning electron microscopy
At both Dryden Flight Center and Kennedy Space Center, initial fixation of cartridge contents with 0.5% glutaraldehyde-2% paraformaldehyde as described above proceeded for ∼2.5 h, after which the fixative was freshly replaced and the cartridges were left an additional 24 h at 4°C. Next, fixative was replaced with 0.05 M cacodylate buffer and the cartridges from both Dryden and Kennedy sites were express-shipped at 4°C to Children's Hospital for further processing. On removal of buffer solution with a syringe connected to an end port, cartridges were opened at both ends with a PVC pipe cutter and the complete contents of capillary fibers, cell-bead mixture, and matrix were placed for 30 minutes in 50-ml Falcon tubes containing 70% ethanol as the first step in dehydration. Subsequent single changes into 80, 95, and 100% ethanol were made every 15 minutes to complete dehydration. Next, fibers with attached cells, beads, and matrix were placed in 100% ethanol in Petri dishes and separated into smaller segments with a razor blade or scalpel. These samples were critical point dried from 100% ethanol using a Denton DCP-1 drying apparatus (Denton Vacuum Co., Cherry Hill, NJ, U.S.A.), sputter-coated with gold in a Polaron model E5150 unit equipped with a thin film monitor (Electron Microscope Sciences, Agawam, MA, U.S.A.), and examined at 25 kV in a JEOL JSM-35 scanning electron microscope (JEOL Co., Peabody, MA, U.S.A.). Photomicrographs were recorded with Polaroid professional 55 positive/negative sheet film (Polaroid Corp., Cambridge, MA, U.S.A.).
Transmission electron microscopy
The same procedure of sample preparation described above for scanning microscopy was utilized through the complete dehydration sequence. At that point, cartridge contents were again divided into smaller samples and placed into 3 × 3 in. polypropylene dishes filled with 50:50 mixtures of absolute alcohol and Spurr resin. Mixtures were changed after 2 h to 25:75 alcohol/Spurr and 2 h later to 100% Spurr for 24 h. Infiltration continued with a change to fresh Spurr resin for an additional 2 days, after which the dishes were placed in a 60°C oven to harden the resin.
The resin embedment was removed from dishes and specimen regions of interest were cut from hardened Spurr. These samples were remounted in appropriate orientations on premade Spurr resin blocks and readied for ultramicrotomy. Thin (∼80 nm) tissue sections were taken with a Reichert Ultracut S (Leica, Deerfield, IL, U.S.A.), collected on 75-mesh grids, stained with uranyl acetate and lead citrate, and photographed in a JEOL 100C (JEOL Co.) or Philips EM300 (Philips Corp., Mahwah, NJ, U.S.A.) electron microscope, each operated at 60 or 80 kV and equipped with a liquid nitrogen anticontaminator.
Measurements of glucose uptake and lactate utilization by cells in respective basal, flight, and ground control cartridges are presented in Table 1, comparing endpoint determinations of the metabolites. As would be expected with metabolically active and viable cells, changes in each metabolite over the time from initial culturing to cartridge retrieval after shuttle landing showed that glucose decreased and lactate increased for both committed and uncommitted cells (data not presented). At the end of the flight, no differences within statistical error in either glucose or lactate were found between any group of cells flown and remaining on the ground (Table 1). Basal cell groups, either committed or uncommitted, terminated at shuttle launch, yielded significantly higher glucose levels compared with either their counterpart flight or ground control cell groups. No statistically significant differences were detected in lactate production among all groups (Table 1).
Table Table 1.. Endpoint (Shuttle Flight Termination) Glucose Residual and Lactate Production Levels Determined from Bone Cells Attached to Cytodex Beads
RNA recovery and analysis
Table 2 presents the measures of total RNA recovered from uncommitted and committed flight and control cells, derived from cultures of 14- and 17-day-old chick embryos, respectively. Statistically significant differences in RNA recovery between flight and control were documented for uncommitted cells (p = 0.010) but not for committed cells (p = 0.353). Figure 1 gives the comparison of collagen and osteocalcin gene expression from the same cartridges. Both type I collagen and osteocalcin genes were expressed by cells in all flight and ground cartridges. Northern blot analysis of the steady-state levels of pro-α1(I) mRNA shows that (1) the mRNA expression of the gene in committed and uncommitted control cell cartridges was equivalent (lanes 1 and 4), (2) the expression of this mRNA in the committed flight cartridge was somewhat reduced compared with that in the uncommitted flight cartridge (lanes 3 and 2, respectively), and (3) the expression of procollagen mRNA was reduced in both committed and uncommitted flight cartridges compared with their respective ground controls (lanes 3 and 1; lanes 2 and 4).
Table Table 2.. Endpoint Recovery of RNA from Bone Cells
The Northern blot analysis for osteocalcin expression shows that (1) the mRNA level in the committed control cell cartridge was greater than that in the uncommitted control (lanes 1 and 4), (2) the level in the committed flight cartridge was greater than that in the uncommitted flight cartridge (lanes 3 and 2, respectively), (3) the level in the committed flight cartridge was less than that in the committed control cartridge (lanes 3 and 1, respectively), and (4) the level in uncommitted flight and control cartridges was equivalent (lanes 2 and 4, respectively).
Figure 1 also shows ethidium bromide staining of total RNA recovered from the cartridges. The well-defined 27S and 18S chicken rRNA bands show that total RNA was intact in both committed and uncommitted flight and control cartridges. Northern blot analysis of 18S rRNA alone defines nearly equivalent loading of gel lanes where the heavy contribution of sample obtained from the ground control cartridge of committed cells is slightly less than the others. For the relative quantitative comparison of expression of type I collagen and osteocalcin in flight and ground control cells, depicted graphically in Fig. 1, any differences in loading were normalized to the respective 18S rRNA. This standardization assumes that rRNA is a constant per cell, and the resulting message expression, given then on a per cell basis, of either collagen or osteocalcin from flight cells is calculated as ∼45–55% that of the control cells, obtained from either 14- or 17-day-old embryos.
Hyrox Hi-Scope light microscopy
Figure 2 illustrates the configuration of a number of hollow fibers, cut transversely and removed from their cartridge housing, and a portion of the cell-bead mixture recovered after spaceflight aboard Endeavour. Beads as well as fiber surfaces serve as scaffolds supporting new cell growth and matrix formation in cartridge bioreactors. Fibers viewed in longitudinal profile show more effectively than transverse outlines the extent of cell and matrix development over the duration of the experiment (Fig. 3).
Scanning electron microscopy
Figures 4 and 5 compare structural features of representative cell-bead mixtures obtained from ground control and flight cartridges. Low-magnification scanning microscopy of two fibers from a control cartridge containing cells derived from 14-day-old chick embryos shows a number of beads and some matrix over one fiber and an extensive bead-matrix development covering the other fiber (Fig. 4A). At increasingly higher magnification (Figs. 4B-4D), this control cartridge, like others examined, is characterized by a thick matrix appearing in sheetlike deposits, themselves obscuring some beads or revealing others at different locations. Some bead surfaces appear with a nearly continuous layer or layers of closely apposing cells with matrix (Fig. 4C), and both of these constituents are plainly identifiable on further enlargement, in which an intricate meshwork of collagen fibrils can be observed as the major component of the extracellular environment (Fig. 4D). The scanning microscopy of a typical flight cartridge containing cells from 14-day-old embryos shows features similar to those found in control cartridges but less extensively elaborated (Figs. 5A-5D). Lower magnification reveals numerous beads and matrix deposition, but the flight cartridge fibers are not covered as prominently with cells and matrix as those recovered from controls (Fig. 5A). Generally round cells can be found on most bead surfaces and matrix is present between beads. The degree of bead coverage by cells or the thickness of matrix is less than that noted in controls (Figs. 5B and 5C). The matrices in flight cartridges, like ground controls, are comprised principally of interwoven collagen fibrils (Fig. 5D).
Transmission electron microscopy
Stained thin sections of representative basal, ground control, and flight cartridge contents typically show poorly preserved and microtomed profiles of Cytodex beads. Distortions in cutting the beads frequently resulted in artifacts of cells and matrix, but occasionally these components were maintained adequately (Figs. 6 and 7). In such instances, cellular ultrastructure could be found consisting of nuclei, endoplasmic reticulum, vacuoles, and other constituents and the matrices were comprised of collagen fibrils. Among the basal, control, and flight samples, whether containing committed or uncommitted cells, it was difficult to determine absolute differences in the number of cells, quantity of matrix, or other structural parameters. On the other hand, transmission microscopy shows that the cellular ultrastructure of ground samples was more intact than that of flight specimens (Figs. 6B-6D), confirms the presence of collagen as a principal extracellular constituent of elaborated matrices in both flight (Figs. 6C and 7C-7D) and control (Fig. 7B) cartridges, and shows normal collagen periodicity (Figs. 7B-7D). Additionally, orthogonal arrays of collagen fibrils are found (Fig. 7C) and mineral deposition is observed in the matrices associated with uncommitted cells maintained either in flight or as ground controls (Figs. 7B and 7D).
Although adaptations of bone to external forces are well documented in studies of the vertebrate skeleton, they are not clearly understood. As noted earlier, they are believed to originate in association with the bone cells, themselves, or their extracellular matrices. That being the case, it is interesting that research is relatively limited with respect to detailed examination of cellular and related responses to skeletal loading. Much of the present data concerning loading has been obtained in clinical or whole animal investigations in vivo.(30–35) The difficulty with cellular approaches, at least in part, appears to be, first, in determining a suitable system for examination of the loading problem; second, in detailing the critical events of cell activity, extracellular matrix production, and mineralization; and, third, in controlling and characterizing the separate loading factors that may possibly act on the cells of interest. The latter would involve cell deformation, hydrodynamic stress (for example, fluid flow, shear, and streaming potential), electromagnetic or electrochemical fluid effects, and additional physicochemical constraints.
While the potential difficulties mentioned above may be formidable, describing skeletal adaptation at the cellular structural level may be accomplished to a considerable degree under certain conditions. One of these is through the use of primary bone cell cultures. Cultures can be established easily, made reproducible, and modified and controlled to allow assessment of specific biological features. In this regard, the primary osteoblast cell culture developed in this laboratory from normal embryonic chicken calvaria and characterized under normal gravity (1G) in many previous studies (Refs. 16–18, for instance) has been flown for the first time in a NASA shuttle mission.
Before a discussion of the results presented above, a number of points should be addressed. First, this experiment was intended to examine spaceflight effects, including those of microgravity, on cultured bone cells. Flight comprises factors besides microgravity that may induce changes in the osteoblasts. These are vibrations on launch of the shuttle, short-duration hypergravity on launch and landing, and radiation exposure throughout the flight time line. Although vibration and hypergravity were of short duration in the spaceflight protocol and radiation also affects ground control cultures, their individual contributions to the nature of the flight cells are unknown. Any possible impact on message expression of the retrieval time of flight cartridges is likewise unknown. The study here could not and did not distinguish the factors and the present data must be considered in this context. Potential vibration and hypergravity effects on bone cultures are being investigated in other work by this laboratory, but information is yet preliminary.
A second point is that the experiment utilized bone cells attached to microcarrier beads as inoculum into cartridges. Unpublished data at the time (E. Holton, personal communication, 1994) indicated that this method of introduction of cells into cartridges was satisfactory for cell growth in the cartridges, but other subsequent work in this laboratory (K. J. Hodgens and W. J. Landis, unpublished results, 1994) showed that cells alone in the bioreactor units were an improvement over the use of beads. Inoculation of cells only was adopted in a second flight protocol a year after that reported here. Regarding the transport of cells between Boston and Kennedy Space Center, they may have been affected to some degree by the absence of CO2 and changes in temperature during the site transition. However, they showed little detectable alteration on arrival at the Space Center. The media always remained bright red/pink in transport, showing that the phenol red indicator was constantly within a normal pH range. More importantly, those cells were replenished immediately in carbonate-buffered media and placed in a 5% CO2 incubator. They were later plated and grown in fresh media under standard conditions and it was these cells that reached the numbers appropriate for cartridge inoculation, approximately the same number of cells being inoculated into all cartridges.
The summary of molecular, biochemical, and structural results carried out here shows that the cells from all bioreactor cartridges were viable and metabolically active over the duration of the experiment but osteoblasts retrieved from flight were distinct from basal and ground control cells in certain aspects. Thus, spaceflight exerted demonstrable effects on the bone cells. In correlates in vivo and in vitro, the general loss of bone mass in response to a diminished mechanical loading or equivalently to the spaceflight-microgravity environment could be the consequence of either or both of two alternatives: a lower cellular activity of mature osteoblasts or a decrease in osteoblast differentiation that produces fewer mature cells. In either case, production of extracellular matrix proteins would be reduced, a situation leading to lower bone mass. In the protocol here, an attempt to address these choices was made by selecting populations of osteoblasts from two different embryonic ages. Populations prepared from 14-day-old embryos were composed principally of cells as yet uncommitted to an osteoblast lineage whereas those isolated from 17-day-old embryos consisted of cells in a more mature stage of their differentiation and thus committed to the lineage compared with the cells obtained from 14-day-old animals.(26, 36, 37) In this context, type I collagen and osteocalcin were then assessed by Northern blot analysis as early (collagen) and late (osteocalcin) markers of osteoblast differentiation (the limited cartridge contents precluded analysis of other markers). Comparative data among the cartridges analyzed for either collagen or osteocalcin appeared consistent in that, first, ethidium bromide staining indicated that respective RNA was intact in the cartridges after the mission, and, second, the osteoblast populations prepared from both 14- and 17-day-old embryos showed progressive differentiation as evidenced by the presence of osteocalcin message.
As might be expected, gene expression for either type I collagen or osteocalcin was greatest in the respective ground control cartridge containing cells derived from 17-day-old embryos, and the expression for both collagen and osteocalcin was again generally higher in all committed control versus uncommitted control and committed flight versus uncommitted flight cartridges for these more differentiated cells compared with their 14-day-old counterparts. An exception to the latter was the collagen expression measured in the committed flight versus uncommitted flight comparison; possibly the relatively higher collagen gene expression here in cells obtained from 14-day-old embryos was the result of an effect on that cell population in response to the longer ascorbate exposure in the prelaunch protocol or of an effect by spaceflight and microgravity in reducing the collagen expression in the cells developed from 17-day-old embryos.
Northern blot inspection further showed that the expression of collagen derived from total cartridge contents was clearly reduced in spaceflight compared with controls regardless of the stage of cell differentiation or duration of ascorbate supplementation normally promoting osteoblast differentiation.(37) Changes in osteocalcin expression after flight were not as readily apparent. With respect to the age of the cell population, Northern blots indicated that, compared with their respective controls, cell cultures developed from 14-day-old embryos were apparently unaffected or only slightly so by flight whereas the cells obtained from 17-day-old embryos yielded a diminished osteocalcin message after flight. This result suggests that the uncommitted cells derived from the 14-day-old embryos do undergo progressive differentiation to mature osteoblasts but committed cells developed from 17-day-old chicks had a slower progression toward differentiation compared with that of their 14-day-old counterparts.
The slower progression to a differentiated function of the cells from 17-day-old embryos may be caused by either a direct regulatory effect by spaceflight on the expression of the differentiated function of these cells or a consequence of lower metabolic consumption by the cells. Regarding the latter, each STL unit provided the same nutritive supply to all cells, although reduced compared with feeding outside the STL, so that any observed changes in their characteristics should not be attributable to nutrition. However, differences in their respective metabolism were difficult to assess here because cell numbers were not determined and changes in endpoint measurements of glucose and lactate in flight and control cartridges in the experiment were not statistically significant.
The subtleties in the appearance of the Northern blots and their implications become considerably more understandable when the amounts of type I collagen and osteocalcin are quantitated relative to 18S rRNA recovered from the respective cartridges. In this case, the message expression of either collagen or osteocalcin from flight cells is only ∼45–55% that of the control cells, regardless of whether the data are derived from either 14- or 17-day-old embryos. This result is presented on a per cell basis given that rRNA contents remain constant per cell.(36, 38–40) The approach obviates a measure of absolute cell number within a given cartridge, a value that was not possible to obtain directly in these experiments. The reduction of osteocalcin expression in flight cells supports the concept that both uncommitted and committed cells follow a slower progression toward a differentiated function than their control counterparts maintained at normal gravity. Stated in another way, although osteocalcin expression is detectable in all cartridges, its reduced expression suggests that spaceflight effects primarily inhibit not the process of commitment and cell differentiation to the osteoblastic lineage but rather the expression of the osteogenic phenotype by the committed osteoblasts.
Although development and changes of the cells in all cartridges of the STL as well as in basal cartridge bioreactors were evidenced by the biochemical and metabolic studies noted above, they also were implicit in comparisons of structure between flight and control cartridges. In this regard, the appearance and elaboration of extracellular matrix, specifically the presence of collagen fibrils about and along the surfaces of the microcarrier beads, were the principal comparative measures. Scanning electron microscopy was most revealing of cell and matrix architecture and, after sampling of recovered cartridge contents, showed greater numbers of cells and more extensive, thicker, and intact collagen layers in control cartridges compared with cells and matrix in the counterpart flight samples. These observations were the same for cartridges containing cells derived from either 14- or 17-day-old embryos and were consistent with the down-regulation of collagen determined in the Northern blots. The observations of many more rounded cells in flight compared with control cartridges regardless of embryo source may be indicative of their response toward sphere formation to optimize surface/volume ratios on exposure to the microgravity environment. An equally plausible explanation is that such shapes are the result of tissue preparation for microscopic examination.
Transmission microscopy, though hindered by some artifacts in sample preparation that prevented quantitative comparisons between cells and matrix recovered from flight and controls, provided unique means of assessing respective ultrastructure. Although less collagen was apparent in flight compared with ground cartridges, the notable presence of orthogonal arrays of the protein was observed during extracellular matrix assembly in flight osteoblasts derived from 14-day-old embryos, a result consistent with matrix formation in vivo.(16, 17) In addition, sites of matrix mineralization were detected in both control and flight cartridges carrying cells obtained from 14-day-old embryos. These findings indicate that the cartridges support cell and matrix development to the important and final point of mineral deposition and that the cells subjected to spaceflight are comparable with control osteoblasts as well as those in vivo in this late event in the mineralization cascade.
The basic result of the present study that spaceflight exerts measurable effects on bone cell cultures is consistent with other investigations of the influence of mechanical forces on bone. In this context, there are numerous references to both loading and unloading cells and animals by a variety of methods but a smaller literature regarding spaceflight. With respect to the latter, Hughes-Fulford and Lewis(41) examined cultured MC3T3-E1 osteoblasts and found that spaceflight altered cell number, nuclei, and stress fibers, but it did not change glucose utilization or prostaglandin E2 synthesis normalized to cell number. Hughes-Fulford and other associates(42–43) also investigated MC3T3-E1cells under loading and vibration simulating a shuttle launch and reported a decrease compared with controls of osteocalcin expression after either the loading or the vibration protocol. Carmeliet et al.(44) has suggested that microgravity decreases the activity of the human osteosarcoma cell line MG-63 (considered representative of osteoblast precursors or early undifferentiated osteoblast-like cells). Their data show a reduction in type I collagen, alkaline phosphatase, and osteocalcin gene expression as well as in cell differentiation in response to systemic hormones and growth factors. Although the decreases in collagen and osteocalcin expression may be related to those reported here, it is difficult to compare data(41–44) because of experimental differences. However, the induction of changes in gene expression by mechanical forces is a common finding.
A similar relation has been documented for bone cells subjected to mechanical strain on deformable surfaces. Stimulation of the same embryonic chick osteoblasts as used in this study has identified osteopontin, integrin receptors, and cytoskeletal elements as critical in mechanoinduction of gene expression in the cells.(38, 45–46) The means for mediating these changes in expression may be related to earlier results showing responsiveness of other osteoblasts to mechanical perturbations(47–49) and to those observed here in type I collagen and osteocalcin under the influence of spaceflight.
Finally, a large number of spaceflight studies have been reported for rats examined as a vertebrate model responsive to mechanical effects. Relevant to the present work, Simmons et al.(50) measured rat mandible bone after an 18.5-day flight and found 30% less mineral and hydroxyproline content in the spaceflight animals compared with controls; Bikle et al.(51) determined that the tibiae from rats subjected to 6 days of flight contained 33% less osteocalcin; and Patterson-Buckendahl et al.(52) showed that rat vertebrae lost 33% of their calcium and humeri suffered 29–35% decreases in mineral content, 25% decreases in osteocalcin, and 40% increases in hydroxyproline after a 7-day flight. On the other hand, Vailas et al.,(53) studying a 12.5-day flight, could not identify changes compared with control animals in rat humerus calcium and phosphorus reflecting mineral content and hydroxypyridinoline and hydroxyproline, markers for collagen. Again, as for the data from cell cultures noted above, these results vary among themselves and among others derived from rats,(54–57) a consequence perhaps of differences in rat age, bone type, flight duration, experimental protocols, and other factors. In any case, it is impractical to compare such data with the collagen and osteocalcin measurements given here although there are some correlates.
This NASA mission was the first in which Cellco Cellmax Quad cartridge bioreactors were utilized to prepare and grow the embryonic chick osteoblasts described above. It was the first that incorporated STL hardware adapted to fly such cells in space, involving a protocol necessarily complex and somewhat uncontrollable in its time lines, limited in its capacity for samples, and requiring careful selectivity of defining analytical assays. Despite these potential difficulties, the experiment showed measurable spaceflight effects on the cultures. In this study, flight data showing collagen and osteocalcin reduced in expression and osteoblasts following a slower progression toward a differentiated function are significant with regard to bone loss suffered by vertebrates, including man, in a weightless environment. The implications of these changes are important with respect to understanding the adaptive character of bone under different loading and unloading regimes and to the function of bone cells under normal gravity. The latter include, for example, conditions of extended immobilization, fracture repair and healing, or exercise and other forms of physical activity. The results obtained here have since been supplemented in a second shuttle flight launched February 3, 1995 (STS-63). Details of this experiment, which support the principal results described here, will be reported separately.
The authors are grateful for the efforts in mission support, discussions, suggestions, enthusiasm, and friendship at both the Kennedy Space Center, FL, and Dryden Flight Center, Edwards Air Force Base, CA, extended by NASA/Ames project scientist, Dr. Emily Holton; Kennedy Space Center STL deputy project manager, Patty Currier; Walter Reed Army Institute of Research (Washington, DC) division director, Dr. William Wiesmann; STL project manager, Tom Cannon; other members of the Walter Reed team, Laura Kerns, Catherine Serke, Ted Delaplaine, Alex Pranger, and Walter Franz; Bionetics Corporation (Kennedy Space Center) team, Mimi Shao (director), Bill McLamb, Grace Spaulding, Brad Berch, Ramona Bober, John Carver, Dennis Turner, and Matt Muller; and NASA/Dryden support team, Joe Calabrese and Darrel Boles. The authors also thank Capt. David Kulesh and SSgt. Bernard Wilson, United States Air Force (Armed Forces Institute of Pathology, Washington, DC), and Dr. Ruth Globus, Alex Malouvier, and Jonathan Lull (NASA/Ames Research Center) for their contributions in helping with this work. Without the assistance of all of the individuals above, this project would not have been possible. The authors also thank Dr. David Zurakowski and James DiCanzio (Children's Hospital, Boston, MA) for assistance with statistical anaysis, Matt Irwin (Electro-Image, Inc., Great Neck, NY) for suggesting the use of a Hyrox Hi-Scope video microscope for imaging intact Cellco Cellmax Quad cartridges and specimens, and William Miller III (Electro-Image, Inc.) for obtaining superior Hyrox photographs. This work was supported by grant NAG-2–538 and NAG-5–7789 from NASA (W.J.L.) and by the Max Kade Foundation and the Austrian Academy of Sciences (C.D.T.).