Compliance of the L5-S1 spinal unit: a comparative study between an unconstrained and a partially constrained system.

A comparison between an unconstrained and a partially constrained system for in vitro biomechanical testing of the L5-S1 spinal unit was conducted. The objective was to compare the compliance and the coupling of the L5-S1 unit measured with an unconstrained and a partially constrained test for the three major physiological motions of the human spine. Very few studies have compared unconstrained and partially constrained testing systems using the same cadaveric functional spinal units (FSUs). Seven human L5-S1 units were therefore tested on both a pneumatic, unconstrained, and a servohydraulic, partially constrained system. Each FSU was tested along three motions: flexion-extension (FE), lateral bending (LB) and axial rotation (AR). The obtained kinematics on both systems is not equivalent, except for the FE case, where both motions are similar. The directions of coupled motions were similar for both tests, but their magnitudes were smaller in the partially constrained configuration. The use of a partially constrained system to characterize LB and AR of the lumbosacral FSU decreased significantly the measured stiffness of the segment. The unconstrained system is today's "gold standard" for the characterization of FSUs. The selected partially constrained method seems also to be an appropriate way to characterize FSUs for specific applications. Care should be taken using the latter method when the coupled motions are important.

Animation of in vitro biomechanical tests.

Interdisciplinary communication of three-dimensional kinematic data arising from in vitro biomechanical tests is challenging. Complex kinematic representations such as the helical axes of motion (HAM) add to the challenge. The difficulty increases further when other quantities (i.e. load or tissue strain data) are combined with the kinematic data. The objectives of this study were to develop a method to graphically replay and animate in vitro biomechanical tests including HAM data. This will allow intuitive interpretation of kinematic and other data independent of the viewer's area of expertise. The value of this method was verified with a biomechanical test investigating load-sharing of the cervical spine. Three 3.0 mm aluminium spheres were glued to each of the two vertebrae from a C2-3 segment of a human cervical spine. Before the biomechanical tests, CT scans were made of the specimen (slice thickness=1.0 mm and slice spacing=1.5 mm). The specimens were subjected to right axial torsion moments (2.0 Nm). Strain rosettes mounted to the anterior surface of the C3 vertebral body and bilaterally beneath the facet joints on C3 were used to estimate the force flow through the specimen. The locations of the aluminium spheres were digitised using a space pointer and the motion analysis system. Kinematics were measured using an optoelectronic motion analysis system. HAMs were calculated to describe the specimen kinematics. The digitised aluminium sphere locations were used to match the CT and biomechanical test data (RMS errors between the CT and experimental points were less than 1.0 mm). The biomechanical tests were "replayed" by animating reconstructed CT models in accordance with the recorded experimental kinematics, using custom software. The animated test replays allowed intuitive analysis of the kinematic data in relation to the strain data. This technique improves the ability of experts from disparate backgrounds to interpret and discuss this type of biomechanical data.

Compressive properties of cancellous bone defects in a rabbit model treated with particles of natural bone mineral and synthetic hydroxyapatite.

A rabbit model was developed to evaluate the compressive mechanical properties of cancellous bone defects treated with particles of selected bone graft substitute materials. A novel feature of the model was the precise retrieval of the site of implantation. A notable finding was a 9-fold increase in the modulus of elasticity of the defect implanted with a synthetic hydroxyapatite material after 26 weeks when compared to the modulus of the trabecular bone normally at the site. The compressive modulus of lesions treated with particles of a natural bovine bone mineral (anorganic bovine bone) was closer to the normal modulus of the cancellous bone at the site. While the compressive strength of the anorganic bone particles was less than that of normal bone, the site implanted with the bone mineral particles achieved compressive strength greater than normal after 6 weeks. Moreover, the anorganic bone particles accelerated the increase in strength of the lesion, at 6 weeks exceeding the strength achieved by the untreated defect after 26 weeks. The potential problem associated with the disparity in the compressive modulus between sites implanted with the synthetic HA particles and surrounding bone is discussed.

In vitro axial preload application during spine flexibility testing: towards reduced apparatus-related artefacts.

Presently, there is little consensus about how, or even if, axial preload should be incorporated in spine flexibility tests in order to simulate the compressive loads naturally present in vivo. Some preload application methods are suspected of producing unwanted "artefact" forces as the specimen rotates and, in doing so, influencing the resulting kinematics. The objective of this study was to quantitatively compare four distinct types of preload which have roots in contemporary experimental practice. The specific quantities compared were the reaction moments and forces resulting at the intervertebral disc and specimen kinematics. The preload types incorporated increasing amounts of caudal constraint on the preload application vector ranging from an unconstrained dead-load arrangement to an apparatus that allowed the vector to follow rotations of the specimen. Six human cadaveric spine segments were tested (1-L1/L2, 3-L2/L3, 1-L3/L4 and 1-L4/L5). Pure moments were applied to the specimens with each of the four different types of compressive preload. Kinematic response was measured using an opto-electronic motion analysis system. A six-axis load cell was used to measure reaction forces and moments. Artefact reaction moments and shear forces were significantly affected by preload application method and magnitude. Unconstrained preload methods produced high artefact moments and low artefact shear forces while more constrained methods did the opposite. A mechanical trade-off is suggested by our results, whereby unwanted moment can only be prevented at the cost of shear force production. When comparing spine flexibility studies, caution should be exercised to ensure preload was applied in a similar manner for all studies. Unwanted moments or forces induced as a result of preload application method may render the comparison of two seemingly similar studies inappropriate.

Serotonin reuptake is less efficient in taste aversion resistant than in taste aversion-prone rats.

We have previously reported the development of rat lines bred selectively for differences in taste aversion conditionability. Earlier studies demonstrated that the taste aversion resistant (TAR) animals exhibited lower concentrations of brain serotonin and consumed greater amounts of ethanol than their taste aversion prone (TAP) counterparts. In the present study, TAR rats demonstrated significantly less efficient brain serotonin transport compared to TAP rats, but the rat lines demonstrated similar levels of serotonin transporter or V(max) and similar whole brain paroxetine (a specific serotonin reuptake inhibitor) binding (B(max)). These results suggest that the rat lines differ in the mechanisms that transport serotonin into nerve endings, but do not differ in the binding of serotonin to the transporter or in the number of serotonin transport sites. The data support the hypothesis that genetically determined differences in the serotonin system contribute to individual differences in taste aversion conditionability. The findings further suggest that differences in serotonin transport may influence the propensity to self-administer ethanol.

Three-dimensional measurement of cemented femoral stem stability: an in vitro cadaver study.

OBJECTIVE: To compare the in vitro stability of two cemented hip stem designs: Stem I was a collarless, double-tapered, highly polished implant; Stem II had a collar and matt finish. BACKGROUND: Stability of the femoral component of a hip implant is important for its long-term clinical success. Excessive migration or cyclic motion can increase the risk of early implant failure. METHODS: The stems were implanted in paired human cadaver femurs, and custom-designed micromotion sensors were used to measure three-dimensional motions of the stems at proximal, middle and distal locations during simulated in vivo loading cycles. RESULTS: This study found that despite 'rigid' fixation, cemented stems exhibit detectable motions under a limited number of cycles of simulated physiologic loads. At four times the donor body weight, Stem I showed a subsidence of 90 microm, compared to 25 microm of Stem II (P<0.05). In contrast, the proximal end of Stem II exhibited greater cyclic motions in the medial-lateral direction (P<0.05). CONCLUSIONS: The different motion patterns could be due to the design differences, such as surface finish and geometry. RelevanceImplant design is an important factor related to the behavior of the cement/bone interface and the overall success of the implant. This study compares in vitro micromotion of two cemented femoral prostheses with differing proximal designs.

Attachment of periosteal grafts to articular cartilage with fibrin sealant.

Favorable results using fibrin sealants in vascular surgery and soft tissue reconstruction have prompted investigation of these biologic adhesives for orthopedic applications. One important recent application was as a sealant for periosteal grafts applied to defects in articular surfaces, a procedure that contained injected chondrocytes cultured in vitro. The low and variable adhesive strength of autologous fibrin substances prompted our investigation of allogeneic fibrin. An in vitro test method was developed to investigate the use of a fibrin sealant for attaching periosteal (bovine) patches to articular cartilage (bovine). Dermis-dermis (porcine) adhesion also was evaluated. In tests of the periosteum-to-cartilage bond performed in a physiological environment, we determined the effects of the following variables on the adhesive shear strength: set time, source of fibrinogen (bovine versus human), and fibrinogen concentration. A specially designed test rig was developed to avoid nonshear force components. Adhesive shear strength increased with fibrin set time for both fibrinogen concentrations and sources (p <.03). The 30-min set time yielded data with less variance than the 5-min set time in all cases except with the higher human fibrinogen concentration (50-80 mg/mL). While there was a trend at each set time towards greater shear strength with increased protein concentration (50-80 mg/mL versus 25-40 mg/mL), only the 5-min trial of the bovine product provided a significant advantage (p <.006). There was no significant difference in adhesive strength between the fibrin products produced with human and bovine fibrinogen. The periosteum-cartilage adhesive strengths obtained in our model were comparable to values recorded for the dermis-dermis bonding. The greater strength at the 30-min set time suggests that a certain time period of joint immobilization might be beneficial in procedures in which grafts are glued to articular cartilage. This study has shown that adhesive strengths achieved with fibrin glues in treating skin wounds also can be achieved in the attachment of periosteal grafts to articular cartilage.

Differences in free-choice ethanol acceptance between taste aversion-prone and taste aversion-resistant rats.

Taste aversion (TA)-prone (TAP) and TA-resistant (TAR) rats were tested for naive, nonforced acceptance of ethanol. Ethanol acceptance had played no role in line development. Rather, the lines had been developed via bidirectional, nonsibling matings based on susceptibility to develop cyclophosphamide-induced conditioned TAs to a 0.1% saccharin solution (at cyclophosphamide doses of 12.5 mg/kg for males and 15.0 mg/kg for females, i.p.). Rats from the 23rd selectively bred generations, with no prior exposure to ethanol, were given 24-hr access to a two-bottle choice between plain tap water and a solution of ethanol in water. Rats were initially given access to 1% ethanol in water, and the ethanol concentration was increased by 1% every 3 days to a maximum of 10%. Ethanol consumption (g ethanol consumed/kg body weight) and preference scores (volume ethanol solution consumed/total fluid intake) were determined by daily bottle weighings. At 1% ethanol concentration, there were no differences between the rat lines in terms of ethanol consumption or preference. At concentrations of 2 to 10%, TAP rats consumed less ethanol and showed a decreased preference for the ethanol solutions than TAR rats. Maximum ethanol consumption was reached at the 6% concentration for both lines. The mean (+/- SE) values of consumption at 6% ethanol were 1.8 (+/- 0.8) and 5.6 (+/- 0.5) g of ethanol/kg body weight for TAP and TAR rats, respectively. Mean (+/- SE) preference scores at 6% ethanol were 26 (+/- 12) and 76 (+/- 6) for TAP and TAR rats, respectively. These findings indicate that differences in TA conditionability may be associated with the propensity of rats to be high or low consumers of ethanol. Based on these data, it is hypothesized that high susceptibility for TA conditionability may deter many individuals from consuming the high levels of ethanol that usually precede alcohol tolerance and dependence.

Factors influencing stability at the interface between a porous surface and cancellous bone: a finite element analysis of a canine in vivo micromotion experiment.

Several factors contribute to the success of stable bony ingrowth into the porous coated surfaces of orthopaedic implants used in hip arthroplasty. Despite having good bony apposition, bony ingrowth might not occur if the relative motion between bone and implant is large. Therefore, determining the limiting micromotion value that inhibits stable bony ingrowth is important. From a previous canine in vivo micromotion study performed at our laboratory, this limiting value was found to be 20 microns. Initially, cementless orthopaedic implants are stabilized only by frictional forces at the bone-implant interface. Therefore, other parameters such as the coefficient of friction and the compressive force normal to the interface should be considered as important factors which stabilize the interface along with micromotion. The purpose of this analytical study was to elucidate how the stability at the bone-implant interface is influenced by various factors, namely, motion of the implant, the coefficient of friction, the degree of pres fit, and the modulus of the surrounding cancellous bone in determining the stability of the bone-implant interface. Nonlinear and linear finite element models which simulated the immediate postsurgical condition and the end point of the canine in vivo micromotion experiment, respectively, were used to this end. From the results of the finite element models it was possible to identify the displacement magnitude for which the implant slipped relative to the bone as the motion of the implant was increased incrementally. This was done for combinations of the coefficient of friction, press fit, and Young's modulus of cancellous bone. This was used as an indicator of the limiting implant motion value beyond which bony ingrowth will be inhibited. The stress distribution in the surrounding cancellous bone bed was also obtained from the results of the finite element analyses for different press-fit conditions. The results of the study indicated that under slight press-fit conditions, the implant slipped relative to bone for implant motions as low as 20 microns. For higher degrees of press fit and reasonable values for the coefficient of friction, no slip occurred for implant motions as much as 100 microns. Although higher degrees of press fit were theoretically conducive to better implant stability, the concomitant high stresses in the adjacent cancellous bone will tend to compromise the integrity of the press fit. This was also evident when the results of an analytical model with a lower degree of press fit correlated well with those of the canine in vivo experiment in which a higher press fit was used, suggesting a possibility of achieving a less than desired press fit during the process of implantation. Through this study the importance of factors other than implant motion was emphasized. The results of the study suggest that the limiting value of implant motion that inhibits bone ingrowth might vary with the degree of press fit for reasonable coefficients of friction.

Factors affecting cement strains near the tip of a cemented femoral component.

A generic three-dimensional finite-element model of the upper half of the femur containing a cemented femoral stem of a total hip arthroplasty was developed to study those factors influencing cement strains near the tip of a cemented femoral component. This generic model was verified through another three-dimensional finite-element model that had been created based on the precise geometry of a cadaver femur implanted with a contemporary cemented femoral component. This cadaveric femoral reconstruction had been created with strain gauges embedded in the cement mantle and was then loaded under conditions simulating single leg stance and stairclimbing. By use of the cement strains measured experimentally in the cadaver femur, and comparison of them with those obtained from the finite-element model of that cadaver femur, it was possible to establish proper material properties, boundary conditions, and loading conditions for the generic model. The generic model was then modified parametrically to determine those factors that influence the strains occurring within the cement mantle near the tip of a cemented femoral component. These models suggest that the single factor that most adversely influenced peak strains at or near the tip of the prosthesis was a thin cement mantle. This effect was present both when the cement mantle was reduced in thickness and when a similar effect occurred by virtue of a varus or valgus placement of the stem. Factors that decreased the peak cement strains near the tip of the femoral stem included a more flexible stem and thicker cement mantles. This effect of a more flexible stem could be obtained by changing the modulus of the metal implant by uniformly reducing the thickness of the stem, or by tapering the stem within the same bone geometry. Thicker cement mantles reduced both the axial and the shear strains occurring at the tip of the prosthesis. The presence or absence of a hole in the tip of the prosthesis per se, as for a centralizer, had no significant effect on the peak cement strains seen around the tip of the prosthesis; however, truncating the tip of the prosthesis from a hemisphere to a flat profile, which resulted in a sharp corner at the tip of the prosthesis, produced a 35% increase in cement strains at the tip as a result of a stress concentration effect. Thus, the common way of modifying the tip to have a hole for a centralizer, which involved truncating the tip, increased the cement strains occurring near the tip of the prosthesis.

Acute immobilization stress disrupts testicular steroidogenesis in adult male rats by inhibiting the activities of 17 alpha-hydroxylase and 17,20-lyase without affecting the binding of LH/hCG receptors.

We have investigated the effect of acute immobilization (3 hours) stress on testicular steroidogenesis in the adult rat. Immobilization did not alter plasma luteizing hormone (LH) levels, but plasma testosterone (T) levels were reduced by 82%. Plasma levels of corticosterone in stressed rats were elevated more than ninefold over control levels. After 3 hours of stress, testicular levels of progesterone were elevated 33%, and levels of 17 alpha-hydroxyprogesterone and T were reduced 47% and 37%, respectively, compared to controls. Immobilization for 3 hours had no effect on the association or dissociation rate constants of LH/human chorionic gonadotropin (hCG) receptors of testicular interstitial cells and did not alter specific hCG binding. The effect of 3 hours of immobilization on testicular 17 alpha-hydroxylase and 17,20-lyase was assessed by incubating testicular microsomes from stressed and control animals in the presence of 21[14C]progesterone and [3H]17 alpha-hydroxyprogesterone. Immobilization of rats reduced the Vmax values of 17 alpha-hydroxylase and 17,20-lyase by 47% and 48%, respectively, but had no effect on the Km values. These results support the hypothesis that stress for 3 hours disrupts rat testicular steroidogenesis via a mechanism that is independent of changes in circulating levels of LH and the binding characteristics of LH/hCG receptors. The effects of immobilization on the content of testicular steroids and on the activities of 17 alpha-hydroxylase and 17,20-lyase suggest that stress inhibits the activities of both 17 alpha-hydroxylase and 17,20-lyase.

Brain levels of amines and amino acids in taste aversion-prone and -resistant rats.

Possible biological contributions to taste aversion (TA) conditionability were explored by comparing whole-brain levels of five neurotransmitter amines and 14 common amino acids within TA-prone (TAP) and TA-resistant (TAR) rats. The selectively bred strains had been developed via 22 generations of bidirectional nonsibling matings based on susceptibility to cyclophosphamide-induced conditioned TAs. The target substances were separated by HPLC and were measured by electrochemical or fluorometric procedures. The TAP brains had higher levels of serotonin (5-HT) and lower levels of norepinephrine (NE) than TAR brains. No strain differences were found with respect to dihydroxyphenylalanine (DOPAC), dopamine (DA), or 5-hydroxyindoleacetic acid (5-HIAA). Among amino acids, TAP rats had lower levels of lysine than TARs: no other differences were detected. Therefore, higher levels of 5-HT and lower levels of NE and lysine were associated with enhanced TA conditionability. The 5-HT and NE results extend prior indications of their central neurotransmitter TA involvements. The functional role of lysine in TA or other brain functions remains obscure.

Skeletal development and bone functional adaptation.

The role of in vivo mechanical loading histories in normal skeletogenesis is related to the process of adaptive, stress-regulated bone remodeling in the adult. The results of many previous computer models for endochondral ossification and bone modeling and remodeling are reviewed. These studies support the view that simple stress-related mathematical algorithms or "construction rules" can be used to emulate normal skeletal development and architectural construction. Such mathematical rules presumably represent the net result of biophysical phenomena influencing cell metabolism and biosynthetic activity. These rules are also successful in describing the adaptation of adult bone to changes in tissue stresses. The findings suggest that stress-related functional adaptation in mature bones may be merely the adult manifestation of the same mechanical construction rules that guide and constrain normal development.

Continued development and unconditioned stimulus characterization of selectively bred lines of taste aversion prone and resistant rats.

This report updates the bidirectional selective breeding of taste aversion (TA) prone (TAP) and TA resistant (TAR) rat lines from the 8th through the 22nd generations. A palatable saccharin solution and the aversive consequences of a cyclophosphamide injection are the respective conditioned stimulus (CS) and unconditioned stimulus (US) of line development. Nonsibling matings within each of the two extremes of TA conditionability have produced TAP and TAR lines having markedly different TA propensities. As previously reported, the substitution of a rotational (i.e., motion sickness) US for cyclophosphamide during TA conditioning also produced characteristic line differences in conditioned taste aversion acquisition. The present report extends the effective line separating USs to include injections of lithium chloride, emetine hydrochloride, and EtOH. A range of EtOH dose levels produced dose-dependent TAs within TAP rats but failed to induce TAs in TAR rats. Following the conclusion of TA testing, the administration of a hypnotic EtOH dose produced equivalent loss of righting capability and equivalent hypothermia in both TAP and TAR rats. The line differences in EtOH induced TA conditionability therefore do not reflect general line differences in EtOH sensitivity. The lines may be useful within studies of biological bases of TA conditionability and animal analog studies of prevention and treatment of alcohol dependence.

Role of glucocorticoids in the stress-induced suppression of testicular steroidogenesis in adult male rats.

We have examined the role of glucocorticoids in the stress-induced inhibition of testicular steroidogenesis. Immobilization (3 hr) reduced plasma testosterone (T) levels to 24% of control values but did not affect plasma LH levels. This reduction was partially reversed by in vivo injections of the antiglucocorticoid, RU486, prior to the stress session at a dose of 10 mg/kg BW, but not at 1.0 or 50 mg/kg BW. Stressed rats that were treated with 10 mg/kg BW RU486 had twofold higher plasma T levels than vehicle-treated stressed animals. Injections of RU486 did not affect plasma LH levels in control or stressed rats and did not affect T levels of unstressed rats. Stressed rats had eightfold higher plasma corticosterone levels than controls, and RU486 had no effect on control or stress levels of corticosterone. The possible role of glucocorticoids in mediating the effect of stress on testicular T production was investigated also in vitro by incubating testicular interstitial cells from unstressed rats for 3 hr with corticosterone (0, 0.01, 0.1, or 1.0 microM) or dexamethasone (0, 0.001, 0.01, or 0.1 microM), followed by an additional 2 hr with hCG (0, 25, 50, or 100 microIU). Both corticosterone and dexamethasone inhibited hCG-stimulated T production in a dose-dependent manner. Cells incubated with the highest concentration of either of the glucocorticoids showed significantly reduced responses to hCG stimulation. In the absence of hCG, in vitro T production was not affected by dexamethasone or 0.01 and 0.1 microM corticosterone. However, the highest dose of corticosterone (1.0 microM) produced a 63% elevation in basal T production. Coincubation of testicular interstitial cells with corticosterone (1.0 microM) or dexamethasone (0.1 microM) and RU486 (0.01, 0.1, and 1.0 microM) reversed the glucocorticoid-induced suppressions of T production in a dose-dependent manner. Our results suggest that during stress increases in plasma levels of glucocorticoids in male rats act via glucocorticoid receptors on testicular interstitial cells to suppress the testicular response to gonadotropins, and that the decline of testosterone production during immobilization stress is in part mediated by a direct action of glucocorticoids on the testis.

Musculoskeletal ontogeny, phylogeny, and functional adaptation.

Physical forces applied to connective tissues may cause significant changes in cell metabolism and gene expression. Theoretical investigations indicate that mechanical loading histories beginning very early in skeletal development may guide endochondral ossification patterns and the initial architectural construction of bones. Developmental patterns and structures of bones can be emulated using mathematical algorithms or "rules of construction" which relate developmental processes to tissue stress (or strain) histories. Skeletal forms and tissues are well-designed for their mechanical function primarily because their histomorphological construction has been guided by mechanical loading during growth and development. Construction rules of developmental mechanics can also be used to describe many of the histological and morphological adaptations of mature skeletal tissues to changes in customary physical activity. Over many generations, changes in the heritable genetic information occurs by mutation and genetic variability. The range of skeletal forms that are possible in evolution due to such variations, however, is constrained by the developmental rules of construction that reflect biophysical processes associated with the tissue mechanical loading.

Computer predictions of bone remodeling around porous-coated implants.

Computer simulations of bone remodeling in response to mechanical stresses can be used to understand normal growth and development of the skeleton or to predict the remodeling of bone in response to prosthetic devices. Using a previously derived bone maintenance theory, a technique for computing bone density distributions was applied to the proximal femur and tibia using two-dimensional, multiple-loading finite element models. The models initially represented solid, homogeneous structures. Using an iterative bone remodeling technique that relates bone apparent density to loading history, the internal distributions of apparent density and elastic modulus for the normal bones were predicted. The finite element models were then modified to represent bones in which porous-coated femoral surface replacements and tibial tray components had been implanted. The same iterative remodeling method was then applied to predict the distribution of bone around these components. The predicted bone density distributions for the natural femur and tibia agree with previously documented normal bone morphology. The predicted bone density distributions around various implanted prostheses were characteristic of the component under investigation and were consistent with clinical and experimental findings of other investigators. In the femoral head, stress shielding occurred underneath the metal surface replacement cup, resulting in lower densities in the femoral head. The addition of a central femoral cup fixation peg caused bone hypertrophy around the peg. In the tibia, the stress concentrations around the pegs also resulted in denser bone, with a concomitant decrease in bone density at more peripheral locations underneath the prosthetic tray. This remodeling technique has the potential to be an important tool in predicting the possible remodeling consequences of new implant design features.

Effects of restraint stress on plasma LH and testosterone concentrations, Leydig cell LH/hCG receptors, and in vitro testicular steroidogenesis in adult rats.

We examined the effect of restraint stress (3 hr) on plasma LH and testosterone levels, on the Leydig cell LH/hCG receptor, and on the activity of enzymes in the testicular steroidogenic pathway of the adult rat. Restraint stress caused a 47% reduction in plasma testosterone concentrations, but had no effect on plasma LH levels. The binding capacity and affinity of Leydig cell LH/hCG receptors were not affected by restraint. Stress did not affect the testicular activity of 20,22 desmolase or 3 beta-hydroxysteroid dehydrogenase, but testicular interstitial cells of stressed rats incubated in vitro with progesterone as a substrate produced more 17 alpha-hydroxyprogesterone but less testosterone than control cells, and when incubated with 17 alpha-hydroxypregnenolone, produced 39% less androstenedione and 40% less testosterone than control cells. These results suggest that restraint stress inhibited 17,20 desmolase but not 17 alpha-hydroxylase activity. When the delta 4 pathway was blocked with cyanoketone (3 beta-HSD inhibitor), stress did not alter the production of pregnenolone or 17 alpha-hydroxypregnenolone, but the production of dehydroepiandrosterone by cells from stressed rats was subnormal, suggesting again a reduction of 17,20 desmolase activity. The data suggest that a major site of the inhibitory action of restraint stress on testicular steroidogenesis is the 17,20 desmolase step. The disruption of androgen production by restraint appears to be LH independent since stress did not affect plasma LH levels, the binding capacity or affinity of LH/hCG receptors, or the activity of 20,22 desmolase.