Search is not available for this dataset
text
stringlengths 1
134k
| id
stringlengths 11
16
| article_id
stringlengths 8
10
| section_title
stringlengths 1
1.35k
| educational_score
float64 0.52
5.16
| domain
stringclasses 3
values | document_type
stringclasses 4
values | domain_scores
listlengths 3
3
| document_type_scores
listlengths 4
4
| language
stringclasses 38
values | language_score
float64 0
1
|
|---|---|---|---|---|---|---|---|---|---|---|
In parallel to producing polyimide, we utilized the lactones (1.3 g, BL/CL = 11/1, mol/mol) derived from the PU tyre to synthesize the biodegradable copolyester, P(BL- co -CL), through a ring-opening copolymerization reaction . Notably, P(BL- co -CL) is a green alternative to petrochemical-based polyolefins , first introduced by Chen et al. . We employed an yttrium complex supported with tetradentate aminoalkoxy-bis-phenolate ligands ( Y-N ), a highly efficient catalyst for the ring-opening polymerization of cyclic esters , for the copolymerization of the relatively ‘nonstrained’ BL and the more ‘strained’ CL. The copolymerization was conducted at −30°C to minimize the effect of the − T Δ S term on the Δ G of the reaction. After 17 hours, the random copolymer P(BL- co -CL) (Mn = 56.3 kg/mol, polydispersity index (Ð) = 1.41) was obtained, with a 73.5% incorporation of BL, alongside a conversion rate of 26.8% for BL and 93.7% for CL . Its thermal properties, analyzed via TGA and differential scanning calorimetry (DSC), revealed a decomposition temperature at 5% weight loss ( T d ) of 226°C , along with a crystallization temperature ( T c ) of −20.9°C and a melting temperature ( T m ) of 17.9°C, characterizing it as a semicrystalline random copolymer .
|
PMC11697979_p15
|
PMC11697979
|
RESULTS AND DISCUSSION
| 4.285059 |
biomedical
|
Study
|
[
0.9994767308235168,
0.00030032245558686554,
0.00022291949426289648
] |
[
0.9992583394050598,
0.00027606089133769274,
0.0003749776224140078,
0.00009064953337656334
] |
en
| 0.999996 |
The synthetic P(BL- co -CL) exhibited excellent mechanical properties, with a toughness of 171 MPa and an impressive elongation at break ( ε B ) of over 1000%, as determined by tensile testing of dog-bone-shaped samples at 5°C . These values surpass those of the corresponding homopolymers, poly (γ-butyrolactone) ( ε B <400%) and poly (ε-caprolactone) ( ε B ≈700%) . The copolymer also displayed an ultimate tensile strength ( σ B ) of 25.0 MPa and Young's modulus ( E ) of 228 MPa , comparable to poly(ethylene terephthalate) and low-density polyethylene . Remarkably, the synthetic P(BL- co -CL) can be fully recycled back to its original monomers (BL and CL) with a near-quantitative conversion rate of over 98% by heating at 250°C for 12.5 hours in the presence of Y(CH 2 SiMe 3 ) 3 (THF) 2 (5 mol%) . The composition of the recycled BL/CL was ∼2.75/1, aligning with the initial composition in P(BL- co -CL). Therefore, the lactones sourced from waste PU tyre have been successfully transformed into P(BL- co -CL) with outstanding chemical recyclability and ductility.
|
PMC11697979_p16
|
PMC11697979
|
RESULTS AND DISCUSSION
| 4.278307 |
biomedical
|
Study
|
[
0.9992126226425171,
0.000302784435916692,
0.00048461381811648607
] |
[
0.9992719292640686,
0.0002491454652044922,
0.0004061155195813626,
0.0000728161830920726
] |
en
| 0.999995 |
In conclusion, this study presents an efficient approach for the upcycling of PU waste, utilizing a novel catalytic process that transforms PU into important chemicals and then valuable polymers. By employing a heterogeneous catalytic system combining methanolysis and hydrogenation, we effectively depolymerized PU into aromatic diamines and lactones in CO 2 /H 2 media. These intermediates were then used to synthesize high-value polymers: polyimide (PI) for advanced engineering applications and polylactone (P(BL- co -CL)) as a biodegradable alternative to traditional plastics. The PI films demonstrated exceptional thermal and dielectric properties, while the synthesized P(BL- co -CL) exhibited remarkable ductility and recyclability. Our approach not only addresses the valorization of PU plastic waste but also offers a sustainable pathway for converting waste into high-performance materials, contributing to a circular economy.
|
PMC11697979_p17
|
PMC11697979
|
CONCLUSION
| 4.150877 |
biomedical
|
Study
|
[
0.9984790682792664,
0.0003222775994800031,
0.001198660465888679
] |
[
0.9990992546081543,
0.0002662168990354985,
0.0005830690497532487,
0.000051473965868353844
] |
en
| 0.999996 |
Insect jumping is actuated by either muscles or springs. For instance, bush-crickets ( Pholidoptera griseoaptera ) use a muscle-actuated system to jump, where the leg muscles contract to extend the legs and this leg extension generates force on the centre of mass . When extending their legs with a muscle contraction, insects are limited in how much mechanical power they can generate by the force–velocity properties of their muscles . By contrast, some insects such as fleas ( Archaeopsyllus erinacei ), froghoppers ( Aphrophora alni ) and grasshoppers ( Schistocerca gregaria ), actuate their jumps by using a spring . These insects initiate a jump by first latching their legs in place and then using large muscles to store energy in a spring-like cuticular structure. After energy is stored in the bending of this cuticular structure, the legs are unlatched and recoil of the spring extends the legs, shooting the animal into the air . This spring actuation allows the animal to circumvent muscle force–velocity trade-offs and thus generate huge amounts of mechanical power, resulting in very high jump speeds . Spring actuated systems using a latch are thus called ‘latch-mediated spring-actuated’ jumpers, or ‘LaMSA’ jumpers for short .
|
39676724_p0
|
39676724
|
INTRODUCTION
| 4.141449 |
biomedical
|
Study
|
[
0.9865773916244507,
0.00044878540211357176,
0.012973818928003311
] |
[
0.7891765832901001,
0.08185805380344391,
0.12831194698810577,
0.0006533515406772494
] |
en
| 0.999996 |
When initiating a jump, a grasshopper first latches its leg by contracting a muscle within its femur, the flexor tibiae. This locks the legs in place . Once the leg is latched, contraction of the extensor tibiae muscle stores elastic energy by deforming the extensor's apodeme and a thickened band of cuticle at the distal end of the tibia called the ‘semi-lunar process’ . The grasshopper then relaxes its flexor tibiae muscle, releasing the latch, causing the semi-lunar process and apodeme to recoil. The recoil forcefully extends the metathoracic leg, rapidly converting the stored elastic energy into kinetic energy. The grasshopper's leg extension pushes into the ground below, accelerating the body and allowing it to become airborne .
|
39676724_p1
|
39676724
|
INTRODUCTION
| 3.343597 |
biomedical
|
Other
|
[
0.7909449338912964,
0.0018458114936947823,
0.2072092592716217
] |
[
0.08871755003929138,
0.9089187383651733,
0.0017290852265432477,
0.000634623400401324
] |
en
| 0.999997 |
Grasshopper kinematics from rigid substrates are well reported in the literature. An adult grasshopper can take off into a jump in an average of 0.025–0.030 s, producing average velocities of 3 to 4 m s −1 with average take-off accelerations of ∼60 m s −2 . Grasshopper take-off elevation angles vary between 28 deg, when the grasshopper moves near horizontally, to 104 deg, when the grasshopper moves near vertically and slightly backwards ; although the average elevation angle is 45 deg . The position of the leg at take-off predicts this elevation . The energetics and power of a jump have also been documented , with a typical adult male grasshopper producing a mean of 9–11 mJ of energy during a jump. Bennet-Clark reported maximum mechanical power of 36 mW. By normalizing this power output by muscle mass, the specific peak power output of the spring recoil is between 450 and 2000 W kg −1 . This is much higher than the 100 W kg −1 of power that can be generated by direct muscle actuation .
|
39676724_p2
|
39676724
|
INTRODUCTION
| 4.113282 |
biomedical
|
Study
|
[
0.9335404634475708,
0.0006025680340826511,
0.06585687398910522
] |
[
0.9919233918190002,
0.0025421413592994213,
0.0054427245631814,
0.00009164032235275954
] |
en
| 0.999996 |
Although these well-known kinematics of grasshopper jumps are all from rigid substrates, grasshoppers sometimes jump from compliant substrates, e.g. from grasses, leaves, or small branches. These are substrates from which applied energy from a jumper can be lost to the movement of the substrate. Kinematics and energetics for S. gregaria for these situations are not currently reported. In other species, the compliant nature of a substrate affects the kinematics and energetics of movement. For example, Cuban tree frogs ( Osteopilus septentrionalis , Hylidae, Anura) can recover energy from a recoiling branch . The orthopteran Locusta migratoria has been shown to have great adaptability to different platform compliances. Across a range of substrate stiffnesses, L. migratoria is able to optimise its take-off velocities, take-off times and accelerations. Should substrates become sufficiently compliant, however, this optimisation is no longer possible .
|
39676724_p3
|
39676724
|
INTRODUCTION
| 3.636655 |
biomedical
|
Study
|
[
0.7509064674377441,
0.0007445446099154651,
0.24834896624088287
] |
[
0.9905445575714111,
0.008495424874126911,
0.0008115615346468985,
0.0001484639651607722
] |
en
| 0.999996 |
Examining this topic synthetically, a 4 g LaMSA jumping robot, inspired by the grasshopper jumping mechanism, was engineered, constructed and jumped from compliant substrates of varying mechanical properties . The experiment investigated the relationship between the mechanical properties, specifically the mass and spring stiffness, of the robotic jumper and the compliant platform. The spring stiffness of the robotic jumper (and will be the same definition when mentioned hereafter) is equal to the stiffness of a theoretical spring that would have the same output energy and the same displacement as the robotic jumper. When the robot jumped from a rigid substrate, most of the energy was expended in the jump and only a little is lost elsewhere, for example, as heat . When jumping from a compliant substrate, however, part of the robot's energy output was lost and contributed to the movement of the platform beneath. The ratio of two key mechanical properties, spring stiffness ( k , in N m −1 ) and mass ( m , in kg), of the compliant platform and the robotic jumper, determined the expenditure of the energy output of the jumping robot. For the robotic LaMSA jumper to recover any lost energy from the recoiling platform, the stiffness of the platform had to be greater than the spring stiffness of the jumper and the mass of the platform had to be less than the mass of the jumper. This created the conditions for the platform to recoil preceding the robot losing contact with the platform, and therefore the energy was able to be returned to the robot as it jumped. However, in the opposite stiffness and mass ratio condition, this energy recovery was not observed because the jumper lost physical contact with the platform prior to the platform's recoil . Whether this principle was applicable to living organisms employing the LaMSA mechanism and jumping from a compliant substrate is unknown.
|
39676724_p4
|
39676724
|
INTRODUCTION
| 4.168897 |
biomedical
|
Study
|
[
0.9857804179191589,
0.0006720963283441961,
0.01354751642793417
] |
[
0.9991841912269592,
0.00045642946497537196,
0.00030509551288560033,
0.00005437980144051835
] |
en
| 0.999996 |
Our study sought to investigate the effects of a compliant platform on the kinematics of LaMSA jumping in grasshoppers. How energy moves between the grasshopper and the substrate, and how this affects kinematics, such as velocity, elevation, acceleration and time to take-off, have not been previously explored and documented. It was hypothesised that grasshoppers fully, or partially, recover energy lost to a compliant substrate, depending on the stiffness and mass of the animal relative to the platform. Therefore, it was predicted that: (1) when the stiffness of the platform was greater than the stiffness of the grasshoppers' legs (herein referred to as ‘grasshopper stiffness’), and (2) the mass of the platform was less than the mass of the grasshopper, the animal would recover lost energy from the recoiling platform, as observed in the robotic LaMSA jumper . In the opposite condition, when platform stiffness was less than the grasshopper stiffness, and when platform mass was greater grasshopper, energy would be lost during a jump and thus compromise kinematics.
|
39676724_p5
|
39676724
|
INTRODUCTION
| 4.108553 |
biomedical
|
Study
|
[
0.9601513147354126,
0.0008788063423708081,
0.03896984085440636
] |
[
0.9993330836296082,
0.0003600431955419481,
0.00025164714315906167,
0.00005520946433534846
] |
en
| 0.999998 |
Fifth instar desert locusts of mixed sex were purchased in batches of 10–12 from a local supplier (Lincoln Reptiles Ltd, UK). Grasshoppers were maintained in a plastic enclosure (27×35×16 cm) within an insectary (ambient temperature maintained at 22–23°C). The enclosure contained egg cartons as shelter and grasshoppers were fed ad libitum with washed cabbage leaves and had continuous access to water in a gel form (HabiStat H 2 O balls).
|
39676724_p6
|
39676724
|
Subjects
| 2.663459 |
biomedical
|
Study
|
[
0.8548547029495239,
0.0010054927552118897,
0.1441398411989212
] |
[
0.8210439085960388,
0.17765440046787262,
0.0008232866530306637,
0.00047838001046329737
] |
en
| 0.999997 |
Active grasshoppers selected from the colony were weighed using an electronic balance (Sarcotorius analytic Avery Weigh-Tronix) before being transferred in a cylindrical plastic sample pot to the testing room. Grasshoppers were left for 30 min to acclimatise to the new environment (a room temperature of 20–21°C) before data collection began. This research was conducted in compliance with ethical standards.
|
39676724_p7
|
39676724
|
Subjects
| 2.14259 |
biomedical
|
Study
|
[
0.7953988313674927,
0.0016744056483730674,
0.20292675495147705
] |
[
0.9331682920455933,
0.06567415595054626,
0.0006674855249002576,
0.0004900945350527763
] |
en
| 0.999996 |
The recording set-up is illustrated in Fig. 1 . Platforms of basswood ( Tilia americana ) of differing sizes (see dimensions defined below), purchased from a local supplier (Hobbycraft, Lincoln, UK), were secured in turn on the top of a rigid wooden box (measuring 13×5×10 cm), that was secured to a table. A 30 cm ruler was fixed to the side of the box to allow for video calibration. All components were secured using re-usable adhesive (Blu-tack ® , it was ensured that this method of adhesion did not add compliance to the system through checks on the films that the platform did not move at the fixed end). The set up was illuminated using a generic floodlight. A high-speed camera (FASTCAM Mini AX with an aspherical LD XR Di SP Tamron AF 28-75 mm F/2.8 F MACRO 67 A09 lens) was positioned approximately 40 cm from, and manually focused on, the platform prior to each video recording. The camera was connected to a laptop computer (Dell G3 15 P75F) used to record the jumps using Photron FASTCAM Viewer recording software (v. 3.6.9.1) set to continuous recording at 1000 frames s −1 and a shutter speed for one frame was 1/10,000 s. The trigger mode set to ‘centre’, which meant that the recording system saved the video footage for 3 s pre- and post- pressing the trigger.
|
39676724_p8
|
39676724
|
Experimental set-up
| 4.101307 |
biomedical
|
Study
|
[
0.9857119917869568,
0.0008723255596123636,
0.013415618799626827
] |
[
0.9973956346511841,
0.0022825547493994236,
0.00023116788361221552,
0.00009063730976777151
] |
en
| 0.999998 |
For control jumps, a basswood platform (0.8×5×61 cm, depth×width×length) was secured to the top of the wooden box, where the fixed portion served as a rigid platform. The two experimental conditions (A and B) used basswood boards that were compliant to varying degrees and were secured to the wooden box in a way so that the compliant portion of the platform could overhang, like a sprung diving board over a pool. The compliant platforms were crafted from pieces of basswood 0.2 cm in thickness and were designed to vary in stiffness and mass, relative to the grasshoppers' stiffness and mass. From here onwards, the relationships between the platform and grasshopper are referred to in relation to the mass ratio ( ) and stiffness ratio ( ), being less than, or greater than, 1.0 (where the stiffness and mass of the grasshopper is the numerator, and the stiffness and mass of the platform is the denominator, see Eqns S1 and S2 in the Supplementary Materials and Methods ). Platform A measured 2.5 cm wide×61 cm long, mass 6.34 g and secured to the end of a box using re-usable adhesive. Platform B was 2.5 cm in width and only 12.4 cm in length with the end portion of this platform, measuring 2 cm, fixed to the supporting box, and at 1.01 g weighed less than the grasshoppers, which averaged 1.13±0.04 g.
|
39676724_p9
|
39676724
|
Experimental set-up
| 4.075342 |
biomedical
|
Study
|
[
0.9469471573829651,
0.0006005861796438694,
0.05245225876569748
] |
[
0.9989112615585327,
0.000828261487185955,
0.0002143647725461051,
0.00004619863102561794
] |
en
| 0.999997 |
The lengths of platforms A and B were divided up into equal parts by lines set 5 cm apart . To contextualise this, the mean grasshopper body length was 3.49±0.04 cm, making the lines 1.43 times longer than the average body length. Platform stiffnesses were manually measured at each of the 5 cm increments by measuring the distance of platform displacement under a calibrated weight of either 1, 2, 10, 20, 50 or 100 g (see Eqn S3 in Supplementary Materials and Methods ), prior to filming the jumping. From here onwards, these marked positions on all compliant platforms will be referred to as ‘line(s)’, where line 1 was the stiffest position on the platform and line 10 was the least stiff and varied along the length of the platforms (see Table S1 ). Differences between the two experimental groups were in the properties of the platform, the number of jumping positions along the platform and the number of jumps per grasshopper.
|
39676724_p10
|
39676724
|
Experimental set-up
| 3.993536 |
biomedical
|
Study
|
[
0.9706320762634277,
0.000529382552485913,
0.028838571161031723
] |
[
0.9995324611663818,
0.00027722030063159764,
0.00015356102085206658,
0.00003680803274619393
] |
en
| 0.999995 |
For control jumps, grasshoppers were placed on the rigid basswood surface and encouraged to jump from right to left by tapping on nearby surfaces to create a sudden sound or by using a small paintbrush to stroke the abdomen. Active, jumping grasshoppers were selected from the enclosure, but if grasshoppers behaved abnormally during the jumps (e.g. jumped using only one metathoracic leg) they were excluded from the analysis. Using vernier callipers (Starrett), each grasshopper's tibia length was measured to an accuracy of 0.1 mm after experimental data collection and before being returned to the enclosure.
|
39676724_p11
|
39676724
|
Experimental procedure
| 3.446142 |
biomedical
|
Study
|
[
0.8108471632003784,
0.0010770413791760802,
0.18807575106620789
] |
[
0.994632363319397,
0.004980064462870359,
0.0002820849767886102,
0.00010553378524491563
] |
en
| 0.999998 |
Three control jumps were recorded from each grasshopper, with 2 min between jumps to allow for rest. Grasshoppers are able to repeatedly jump without measurable fatigue (up to 61 consecutive jumps) , although a 2 min rest was implemented for the additional use of allowing time to process the films between jumps. The sample size of grasshoppers and total number of jumps used for control jump analysis are shown in Table S1 . After control jumps were recorded, grasshoppers were jumped from each line along one of the two platforms, totalling 10 experimental jumps from each individual for platform A and two experimental jumps for the shorter platform B with 2 min rests between each jump. Each grasshopper jumped from each line in a random order and were jumped sequentially allowing for a 2 min recovery period between jumps. The grasshopper was placed on the platform so that their centre of mass was between the marked line and 2 cm behind the line for each jump.
|
39676724_p12
|
39676724
|
Experimental procedure
| 3.887638 |
biomedical
|
Study
|
[
0.9346891641616821,
0.0007608993910253048,
0.0645499899983406
] |
[
0.9990335702896118,
0.0007509153801947832,
0.00016304737073369324,
0.00005243551640887745
] |
en
| 0.999997 |
Videos were digitized using Tracker Video Analysis and Modelling Tool . Each video was calibrated to scale using the 30 cm ruler in camera view . Initial measurements were made by tracking the metathoracic leg joint or on the caudal–ventral extremity end of the pronotum, depending on which area was continuously visible throughout the footage. Positions were measured at the last frame of tarsi-platform contact and 10 ms later .
|
39676724_p13
|
39676724
|
Experimental procedure
| 3.882519 |
biomedical
|
Study
|
[
0.9977455735206604,
0.0003455300466157496,
0.001908912556245923
] |
[
0.9974285960197449,
0.0022114457096904516,
0.0002704224025364965,
0.00008950591291068122
] |
en
| 0.999996 |
For experimental jumps only, a second set of plots were placed in the software beneath the left tarsus of the grasshopper in the first frame of visible metathoracic movement and the second being the last frame of tarsi-platform contact. These plot coordinates were used for the measurement of platform displacement (m) during the grasshopper jump. An in-software protractor allowed measurement of the angle of metathoracic leg extension for control jumps to calculate the grasshopper's acceleration distance ( Eqn S4 ) and stiffness ( Eqn S5 in Supplementary Materials and Methods ).
|
39676724_p14
|
39676724
|
Experimental procedure
| 4.003531 |
biomedical
|
Study
|
[
0.973861038684845,
0.0004800963797606528,
0.02565896324813366
] |
[
0.9990350008010864,
0.0007223497377708554,
0.0001945062685990706,
0.000048166544729610905
] |
en
| 0.999998 |
The stiffness of grasshoppers during a jump ( k g ) was calculated by determining the stiffness of the linear spring that would produce the same kinetic energy as that observed during a jump from a stiff substrate (the control jumps for each platform). This was done by applying conservation of energy: (1) where m g is the mass of the grasshopper, v is the take-off velocity and x is the distance the grasshopper accelerates before take-off; x was calculated by determining the distance over which the centre of mass accelerated during the extension of the legs. Gravity has not been considered in the energy calculations because only 1–2% of the grasshopper energy budget is spent on gravity at these masses and distances . At the beginning of the jump, the legs were fully flexed (with the distal end of the tibia and the proximal end of the femur almost touching). At take-off, the legs were fully extended, allowing the acceleration distance ( x ) to be calculated in terms of the length of the femur ( L f ) and the tibia ( L t ) with the law of cosines: (2) where α is the angle of the femur-tibia joint at take-off, and L f and L t mark the sides of an isosceles triangle . Eqn 2 was then be substituted into Eqn 1 to evaluate the effective stiffness ( k g ) of the grasshopper legs during a jump.
|
39676724_p15
|
39676724
|
Analysis
| 4.16256 |
biomedical
|
Study
|
[
0.9751201868057251,
0.0006075194105505943,
0.024272359907627106
] |
[
0.998897910118103,
0.0007709676865488291,
0.00027408290770836174,
0.00005699614484910853
] |
en
| 0.999998 |
High-speed movies were used to calculate the following variables using equations detailed in the Supplementary Materials and Methods ( Table S2 ): time to take-off ( Eqn S6 ), velocity ( Eqn S7 ), elevation relative to the deformed platform plane ( Eqn S8 ), platform displacement angle ( Eqn S9 ), acceleration distance ( Eqn S4 ), acceleration ( Eqn S10 ), kinetic energy ( Eqn S11 ), kinetic energy density ( Eqn S12 ), power ( Eqn S13 ), power density ( Eqn S14 ), grasshopper stiffness ( Eqn S5 ), platform stiffness ( Eqn S3 ), stiffness ratio between the platform and the grasshopper ( Eqn S1 ), and mass ratio between the platform and the grasshopper ( Eqn S2 ). Prior to carrying out statistical analysis, a mean of each variable for each grasshopper's set of control jumps was calculated [(variable) c ]. This was used to normalize experimental values relative to the mean control value of each variable. For all variables proportions relative to the mean of the control was calculated ( Eqn S15 ), except for elevation where difference was used ( Eqn S16 ). Difference was used only for elevation because this variable is a polar quantity and therefore a scalar statistic could not be used.
|
39676724_p16
|
39676724
|
Analysis
| 4.116632 |
biomedical
|
Study
|
[
0.9915346503257751,
0.00044911596341989934,
0.008016259409487247
] |
[
0.9994958639144897,
0.00031281838892027736,
0.00014969092444516718,
0.000041599552787374705
] |
en
| 0.999995 |
For control jumps, statistical analysis used R packages lme4 and lmertest to perform linear mixed-effects models in R ( R-project.org ). For experimental jumps, the mean k g was 3.42 N m −1 , whereas k p varied according to the position of the line and the platform (see Table S1 ). The data were split into two categories based on stiffnesses of the platform relative to that of the grasshopper: i.e. k p < k g or k p > k g .
|
39676724_p17
|
39676724
|
Statistical analysis
| 3.877077 |
biomedical
|
Study
|
[
0.9840360283851624,
0.0003182519576512277,
0.015645762905478477
] |
[
0.9987919926643372,
0.0009427045006304979,
0.00021208867838140577,
0.00005327580583980307
] |
en
| 0.999996 |
Models tested either the relationship between platform stiffness as a covariate and the different variables for when stiffness of the platform was less than that of the grasshopper, i.e. k p < k g . When stiffness of the platform was greater than that of the grasshopper, i.e. k p > k g , the models tested the effect of platform stiffness as a covariate, but also included platform type as a fixed factor, for the different variables. These latter models initially included an interaction term, which if it proved non-significant, was removed and the model was re-run. All models included a code for each grasshopper as a random factor to control for repeated sampling.
|
39676724_p18
|
39676724
|
Statistical analysis
| 3.563804 |
biomedical
|
Study
|
[
0.8840717077255249,
0.0009527353104203939,
0.11497559398412704
] |
[
0.9965674877166748,
0.0030348924919962883,
0.000295870762784034,
0.00010178445518249646
] |
en
| 0.999996 |
A one-sample t -test was used to test whether mean for each variable at each given platform stiffness was significantly different from a mean of 1.0. The exception was elevation, which was tested against a mean of zero because comparison of the control and experimental jumps was a mean difference rather than a mean proportion. All datasets were tested for a normal distribution using a Shapiro test. All but six datasets were normally distributed, and when tested using a Wilcoxon test, produced the same result as the t -test.
|
39676724_p19
|
39676724
|
Statistical analysis
| 4.051946 |
biomedical
|
Study
|
[
0.998095691204071,
0.0002774468739517033,
0.0016268388135358691
] |
[
0.999100923538208,
0.0006330757751129568,
0.00021486870537046343,
0.00005116673855809495
] |
en
| 0.999998 |
The 40 grasshoppers used for this study had a mass of 1.13±0.04 g (mean±s.e.m.) and had jumps from a fixed substrate that took off with a velocity of 1.44±0.03 m s −1 . The kinetic energy density was 24.79±0.93 J kg −1 . The length of the tibiae was 15.6±0.187 mm and angle of extension of the femur/tibia joint at take-off was 113±2 deg. The grasshoppers jumped as if propelled by a spring of stiffness ( k g ) of 3.39±0.11 N m −1 (see Eqns 1 and 2 for how grasshopper stiffness was calculated). There were no significant differences for the different variables between the control jumps for the two groups of grasshoppers used for the two platforms ( Table S3 ).
|
39676724_p20
|
39676724
|
Effective stiffness of grasshoppers during jumps
| 4.133859 |
biomedical
|
Study
|
[
0.9961275458335876,
0.0004402415652293712,
0.003432255471125245
] |
[
0.9996179342269897,
0.00020144747395534068,
0.00013508791744243354,
0.00004544534385786392
] |
en
| 0.999998 |
This analysis was based on data collected from platform A only, as this condition occurred when the stiffness of the platform ( k p ) was less than the mean stiffness of the grasshoppers ( k g ). Time to take-off was significantly positively affected by platform stiffness . By contrast, elevation was unaffected by platform stiffness .
|
39676724_p21
|
39676724
|
Kinematics of jumps when platform stiffness is less than mean grasshopper stiffness
| 2.278029 |
biomedical
|
Study
|
[
0.8440423607826233,
0.0016287543112412095,
0.15432879328727722
] |
[
0.9924033880233765,
0.007059421855956316,
0.00036208046367391944,
0.0001750754745444283
] |
en
| 0.999998 |
There was a significant positive relationship between k p and relative velocity ( ) . Likewise, mean values for kinetic energy density relative to control KED ( ) exhibited a strong positive relationship between k p and , but only until k p reached the mean k g . Although there was a slight positive relationship between k p and relative acceleration this was non-significant ( Table 1 ). When k p < k g , like the pattern for acceleration, the relative power density increased with platform stiffness but this relationship only approached significance at 0.05 ( Table 1 ). Plots for the relationships between platform stiffness and kinetic energy and power are shown in Fig. S2C,D alongside the equivalent analysis ( Table S4 ).
|
39676724_p22
|
39676724
|
Kinematics of jumps when platform stiffness is less than mean grasshopper stiffness
| 4.053895 |
biomedical
|
Study
|
[
0.9861424565315247,
0.0004075951292179525,
0.013449987396597862
] |
[
0.9996281862258911,
0.00018900373834185302,
0.00014756801829207689,
0.00003528272645780817
] |
en
| 0.999996 |
There was no significant interaction between time to take-off and platform type . In the simplified model, neither platform stiffness nor type significantly affected time to take-off ( Table 1 ). Similarly, elevation was unaffected by the interaction between platform stiffness and type and neither the covariate nor fixed factor were significant in the simplified model ( Table 1 ).
|
39676724_p23
|
39676724
|
Kinematics of jumps when platform stiffness is greater than mean grasshopper stiffness
| 2.365908 |
biomedical
|
Study
|
[
0.9178916215896606,
0.0017867175629362464,
0.08032165467739105
] |
[
0.9970287680625916,
0.0025008099619299173,
0.0003195765893906355,
0.00015087006613612175
] |
en
| 0.999996 |
There was no obvious effect of platform stiffness or platform type on take-off velocity . Linear mixed effect regression modelling showed no significant interaction between k p and platform type for relative velocity ( F 2,120.72 =0.92, P =0.402). In the simplified model there was no significant effect of k p on ( Table 1 ). Mean values for kinetic energy density relative to were not significantly affected by platform stiffness and the trendlines had intercepts that were close to the value of 1.0 . Linear mixed effects regression modelling found no interaction was found between k p and platform type ( F 2,112.30 =0.618, P =0.541). There was no significant effect of platform stiffness or platform type on in the simplified model ( Table 1 ).
|
39676724_p24
|
39676724
|
Kinematics of jumps when platform stiffness is greater than mean grasshopper stiffness
| 4.059013 |
biomedical
|
Study
|
[
0.9542629718780518,
0.0007040223572403193,
0.0450330413877964
] |
[
0.99946528673172,
0.0003001678560394794,
0.00019071879796683788,
0.000043750726035796106
] |
en
| 0.999997 |
For relative acceleration values for all platforms exhibited slight negative relationships with k p . Linear mixed effects analysis showed no interaction between k p and platform type on relative acceleration ( F 2,123.81 =0.671, P =0.512) and neither platform type nor stiffness showed any significant relationship in the simplified model ( Table 1 ). Mean values for relative power density exhibited slight negative relationships between and k p for both platforms but were not significantly below 1.0 for either platform A or B . Linear mixed effects regression modelling showed no significant interaction between platform type and k p ( F 2,122.96 =0.796, P =0.454) and the simplified model showed no effect of k p or platform on ( Table 1 ).
|
39676724_p25
|
39676724
|
Kinematics of jumps when platform stiffness is greater than mean grasshopper stiffness
| 4.121265 |
biomedical
|
Study
|
[
0.980999231338501,
0.0004956646007485688,
0.01850518025457859
] |
[
0.9995455145835876,
0.0002303944347659126,
0.0001875327725429088,
0.00003652761006378569
] |
en
| 0.999995 |
The importance of the platform stiffness on grasshopper jumping kinematics has been highlighted by the present study. We have found here that grasshopper jump kinematics were affected by a substrate only when the substrate stiffness was less than the stiffness of the grasshopper leg. If the stiffness of the substrate was greater than the stiffness of the leg, then grasshopper kinematics were unaffected. The effect of stiffness was observed on the large platform, but this effect could be ameliorated by platform mass , such that if a platform were heavy enough, then even very low platform stiffnesses would not affect the jump. Only having two mass conditions, we do not have a large enough range of platform masses available to determine the quantitative effect of mass. As pointed out by Divi et al. , for a sufficiently large substrate mass, the stiffness of both the jumper and the substrate become unimportant as the governing mechanics of the jump become a simple exchange of momentum ( m p × v p = m g ×v g ). If the platform is sufficiently compliant and light enough (as the platforms in this experiment were) then take-off velocities, kinetic energy density and power density decrease as platform stiffness decreases. If a substrate is sufficiently light and stiff, however, energy lost to the substrate can be recovered by the grasshopper, as the substrate recoils .
|
39676724_p26
|
39676724
|
DISCUSSION
| 4.070464 |
biomedical
|
Study
|
[
0.9175505638122559,
0.000961451733019203,
0.08148796111345291
] |
[
0.9991021156311035,
0.0005011034081690013,
0.00034125111415050924,
0.00005555433381232433
] |
en
| 0.999997 |
Grasshoppers were repeatedly jumped and although they were given time to recover, there was a possibility for muscle fatigue to occur which could compromise jump performance. Bennet-Clark showed that grasshoppers were fatigued after 5 to 10 jumps, and jump velocity decreased dramatically within the first 5 min of jumping for second, fourth, sixth instars and adults when repeatedly jumped . Whilst fatigue could have been a confounding variable in this study, the control jumps were filmed prior to the experimental jumps, which ensured that each individual had an optimum jump recorded as a baseline. The experimental jumps were also recorded in a randomised order of stiffness, with each grasshopper having rest time between jumps. Moreover, if fatigue was an issue, a decrease in velocity would be observed for the stiff substrate as well. Although muscle fatigue from repeated jumping would decrease velocity and energy, both variables were maintained under high and low conditions. Moreover, Goode and Sutton demonstrated that grasshoppers were able to jump over 60 times without measurable fatigue.
|
39676724_p27
|
39676724
|
Potential confounding variables
| 4.122516 |
biomedical
|
Study
|
[
0.9947109222412109,
0.0003381221613381058,
0.004951042588800192
] |
[
0.9989253878593445,
0.00028128581470809877,
0.00074225157732144,
0.00005106299795443192
] |
en
| 0.999996 |
Resonance frequency of the platform and the jumper have previously been reported in similar studies, such as Divi et al. , who reported that a faster resonance frequency of the platform, relative to the jumper, being key to measuring energy recovery. There remains much to explore about relative frequency, although we decided to not include this in our study because of the large amount of data focusing on kinematics already being reported. Further investigation is necessary to extrapolate exactly how much energy is being recovered in biological LaMSA systems while keeping resonance frequency in mind.
|
39676724_p28
|
39676724
|
Potential confounding variables
| 2.385981 |
biomedical
|
Study
|
[
0.8014727234840393,
0.0007541858940385282,
0.1977730095386505
] |
[
0.9781198501586914,
0.020759157836437225,
0.0009068271028809249,
0.00021419751283247024
] |
en
| 0.999997 |
The kinematics of grasshoppers jumping from rigid platforms reported here was comparable to data reported in the literature. For instance, adult grasshoppers jumped at 3.2 m s −1 , which was faster than earlier ontogenetic stages . In this study, fifth instar grasshoppers jumped at a mean velocity of 1.45 m s −1 , which was slower than an average take-off velocity of 1.5 m s −1 for fourth instar juvenile grasshoppers reported by Katz and Gosline . These differences may reflect increases in cross section area of the femur which becomes wider through moults to accommodate the increase in the mass of the extensor tibiae muscles .
|
39676724_p29
|
39676724
|
Performance of grasshoppers in control jumps
| 4.031322 |
biomedical
|
Study
|
[
0.9672276973724365,
0.0005177700077183545,
0.03225449100136757
] |
[
0.9991808533668518,
0.0005727859097532928,
0.0002000854874495417,
0.00004620867548510432
] |
en
| 0.999997 |
Jump elevation is independently controlled by the rotation at the coxae, prior to tibia extension . In the present study, the insect jumped with a mean elevation angle of 43.5 deg ( Table S3 ), which was comparable to adult grasshoppers that consistently jumped at 45 deg . Grasshoppers can jump at elevations ranging from 28 to 104 deg but, despite this, the fifth instars jumped here were very consistent. Visual input and a ‘peering’ behaviour prior to the jump allow the grasshopper to jump accurate distances and jump towards a specific target that requires a precise jump elevation . The combination of a consistent jumping environment, and the independent control of the jump elevation by rotation at the coxae was likely to have created this reliable mean elevation of 43 deg. However, jump elevations in the present study ranged from 19 to 69 deg, which may reflect different jump types . For instance, ‘directed’ jumps are preceded by the peering or scanning behaviour, but this is not often observed during ‘escape’ jumps . Though not systematically recorded in this study, this peering behaviour was occasionally observed prior to jumping by some individuals. Grasshoppers that were motivated to jump may be expressing an escape jump that lacked the peering or scanning behaviour and rotation of the coxae and so may explain a wide range of observed elevations in this study.
|
39676724_p30
|
39676724
|
Performance of grasshoppers in control jumps
| 4.180151 |
biomedical
|
Study
|
[
0.9961552023887634,
0.0004232722276356071,
0.0034215429332107306
] |
[
0.9995717406272888,
0.0002357504708925262,
0.00014391964941751212,
0.00004853766586165875
] |
en
| 0.999997 |
Literature around locomotion from compliant platforms is limited and none are available for S. gregaria . Orangutans ( Pongo abelii ) can maximise their energy budget from the oscillating movements of branches while swinging from tree to tree . When jumping Cuban tree frogs recover energy from compliant branches, even from a static position, velocity, energy and power output of the frogs are unaffected . This is achieved by delaying the extension of the hind leg, which increased the time for which the frog is in contact with the substrate . By contrast, compliant perches significantly decrease jump velocity in green anole lizards . Mo et al. found that Locusta migratoria were able to jump optimally on complaint ground in relation to their time to take-off, velocity and acceleration.
|
39676724_p31
|
39676724
|
Jumping performance from compliant platforms
| 4.008624 |
biomedical
|
Study
|
[
0.9925991296768188,
0.00020210316870361567,
0.00719874445348978
] |
[
0.9954032897949219,
0.0017223391914740205,
0.0027931940276175737,
0.00008116824756143615
] |
en
| 0.999996 |
Grasshoppers usually locomote across compliant vegetation where the ability to recover energy when jumping could be advantageous. It is possible that grasshoppers select where they jump from to optimise energy recovery, therefore enhancing jump performance in the field. This would require perception of mass and stiffness of their environment, and the cognitive ability to perceive the environment. In addition, one recent study has revealed that grasshoppers can filter out some sensory information from visual input, supporting the idea that they have a developed cognitive ability to make informed choices about their environment .
|
39676724_p32
|
39676724
|
Jumping performance from compliant platforms
| 1.840963 |
other
|
Study
|
[
0.2310517132282257,
0.001220892765559256,
0.7677273750305176
] |
[
0.5132741928100586,
0.48228615522384644,
0.0033713674638420343,
0.0010683228028938174
] |
en
| 0.999999 |
Interestingly, a model robot, inspired by the grasshopper's LaMSA system, can recover energy from a recoiling platform under specific mechanical conditions . When the platform stiffness was greater than the robot stiffness, energy recovery was observed. When replicating this methodology for the investigation into grasshopper LaMSA jumping, there was a stress on the importance of relationship between the mass and stiffness of the platform and grasshopper leg. Across all ratios of masses and stiffnesses between the grasshopper and the platform, there was no significant influence of the platforms on time to take-off and elevation but there were significant relationships for velocity and kinetic energy density but non-significant relationships for acceleration and power density as proportions of the control values.
|
39676724_p33
|
39676724
|
Jumping performance from compliant platforms
| 3.452725 |
biomedical
|
Study
|
[
0.66450434923172,
0.0008665377390570939,
0.3346291184425354
] |
[
0.9969674944877625,
0.0025445017963647842,
0.00039632857078686357,
0.00009165174560621381
] |
en
| 0.999996 |
Adult grasshoppers take around 0.02–0.03 s to take-off, whereas fifth instars can take up to 0.65 s from a rigid substrate . Experimental times to take off were not greatly different to the control times to take-off and were unaffected by platform mass or platform stiffness. Astley et al. also found that jump duration by Cuban tree frogs did not vary across differing perch stiffnesses and suggested that the frogs have adapted a robust jumping strategy by maintaining their time to take-off while jumping from various perches in the field. This explanation could apply to grasshopper jumping because to produce the same velocities over a slower time to take-off would create lower accelerations and lower power outputs and would therefore be advantageous to keep a constant time to take-off to continue producing high accelerations.
|
39676724_p34
|
39676724
|
What is not affected by platform compliance?
| 3.997505 |
biomedical
|
Study
|
[
0.9416462779045105,
0.00046875415137037635,
0.05788490176200867
] |
[
0.9965762495994568,
0.0023068904411047697,
0.0010331009980291128,
0.00008379712380701676
] |
en
| 0.999996 |
The differences in elevation compared with the controls were non-significant, suggesting that grasshoppers were able to adapt to maintain their elevations across different compliant substrates of differing masses and stiffnesses. Grasshoppers can maintain precise elevations while jumping, through peering and scanning behaviours and rotation at the coxae , which is supported by the precise elevations from the fifth instar control jumps. Consistent jump elevation also suggests that, no matter the reaction of the substrate, the direction of the force vector of the jump stays constant, allowing recovered energy to apply force in the direction of the jump, instead of knocking a grasshopper off target.
|
39676724_p35
|
39676724
|
What is not affected by platform compliance?
| 3.671211 |
biomedical
|
Study
|
[
0.931790292263031,
0.0005470244796015322,
0.06766263395547867
] |
[
0.9972947239875793,
0.0024336720816791058,
0.00020166512695141137,
0.00007001202902756631
] |
en
| 0.999998 |
The patterns of the relationships between platform stiffness and velocity, kinetic energy and kinetic energy density were similar because they were all calculated based on a similar set of kinematic measurements. Velocity is required to calculate kinetic energy and so changes to velocity were squared in the calculation of kinetic energy. By contrast, changes in mass of the grasshopper would affect kinetic energy in a linear fashion ( Eqns S2 and S7 ), although the mass of each grasshopper was constant during this experiment, and therefore did not affect the kinetic energy budget. A reduction in the grasshopper's energy budget only occurred when the platform had a stiffness less than that of the grasshopper leg. Divi et al. also found that the platform from which the LaMSA robot jumped from could sometimes recoil before the jumper lost contact with it. While still in physical contact with the platform, the jumping robot was able to recover up to 12 mJ of this lost kinetic energy back into the jump . This was also observed in grasshoppers because when jumped from a platform with low mass and a high stiffness relative to the grasshopper mass and stiffness ( >1 and <1), the grasshopper maintained its kinetic energy without dissipation to the platform, which resulted in a maintained velocity relative to the control jumps. Energy recovery was possible from a light and stiff platform, relative to the jumper, because it can recoil fast enough for the jumper to recover energy from it. This was observed in stills of the high-speed films of grasshoppers jumping .
|
39676724_p36
|
39676724
|
Connections between velocity and kinetic energy density
| 4.108722 |
biomedical
|
Study
|
[
0.9769135117530823,
0.000660507648717612,
0.022425904870033264
] |
[
0.9993194341659546,
0.0003155407030135393,
0.0003137959574814886,
0.00005124000017531216
] |
en
| 0.999996 |
Grasshoppers were able to recover energy from a platform of high mass and high stiffness, rather than exclusively from low masses as seen in robotic LaMSA systems . Here, the platform does not necessarily need to be a lower mass than the jumper for energy recovery to occur in biological LaMSA systems. However, there is a point in which the platform became too massive for it to recoil quickly enough for energy recovery to occur. Energy recovery has been demonstrated in grasshopper jumping if <1, and >0.17. This suggests that grasshoppers, and possibly other biological LaMSA jumpers, are more robustly adapted to jump from a wider range of substrate masses than initially hypothesized. However, it is imperative that the stiffness of the platform is greater than that of the grasshopper for it to recoil quickly enough. Platform recoiling was also observable in the high-speed film stills of jumps from the smallest platform (an >1 and <1 condition), but energy recovery from a platform where <1 and <1 condition (i.e. the stiff end of platform B) could not be observed from the film stills.
|
39676724_p37
|
39676724
|
Connections between velocity and kinetic energy density
| 3.850427 |
biomedical
|
Study
|
[
0.8497823476791382,
0.0008657728903926909,
0.14935192465782166
] |
[
0.9964296221733093,
0.002928685862571001,
0.0005405863048508763,
0.00010106953413924202
] |
en
| 0.999995 |
For kinetic energy density and velocity to be maintained in nature, the grasshopper would need to select leaves and grasses that are stiffer than their leg stiffness. However, the substrate does not necessarily need to be lighter than the grasshopper, because energy recovery had been demonstrated from platforms that are of greater mass than the grasshopper. Although these conditions are what we could define as optimum jumping conditions, it is unknown whether grasshoppers have the decision-making ability to select for these conditions.
|
39676724_p38
|
39676724
|
Connections between velocity and kinetic energy density
| 2.047592 |
other
|
Study
|
[
0.42063719034194946,
0.0012012567603960633,
0.5781615376472473
] |
[
0.730176568031311,
0.26743045449256897,
0.0015723364194855094,
0.0008205779013223946
] |
en
| 0.999997 |
Acceleration is equal to the velocity at take-off over time, which is required for calculation of power, and power density is power relative to the mass of the grasshoppers' jumping muscles. Consequently, acceleration, power and power density all followed similar trends relative to platform stiffness.
|
39676724_p39
|
39676724
|
Connections between acceleration and power density
| 2.212174 |
biomedical
|
Other
|
[
0.7458248138427734,
0.0010440887417644262,
0.25313109159469604
] |
[
0.42192015051841736,
0.5756359696388245,
0.0016310468781739473,
0.0008128446061164141
] |
en
| 0.999997 |
Even though Divi et al. did not discuss acceleration and power density of the LaMSA robot, it can be hypothesized that these variables would be lower than the control values when the platform mass was high because of the relationship of these variables with velocity and kinetic energy. Data supported this because acceleration and power density of jumps from the smallest platform were similar to the control accelerations, and and remained ∼1.0 across all platform stiffnesses but were much lower on platform A where the mass was much greater than the grasshopper.
|
39676724_p40
|
39676724
|
Connections between acceleration and power density
| 3.261851 |
biomedical
|
Study
|
[
0.8027560710906982,
0.0006882768939249218,
0.1965557336807251
] |
[
0.9965954422950745,
0.003020207630470395,
0.00029665883630514145,
0.00008769395208219066
] |
en
| 0.999998 |
Similarly to velocity and kinetic energy density, acceleration and power density under the conditions on platform A, were close to 1.0, explaining the maintenance of acceleration and power density. However, acceleration and power density did not vary along the different stiffness of platform A, suggesting they were unaffected by stiffness. Similarly, Astley et al. found that the power output of jumping Cuban tree frogs was not significantly affected by branch stiffness, but their energy output was.
|
39676724_p41
|
39676724
|
Connections between acceleration and power density
| 3.157283 |
biomedical
|
Study
|
[
0.8070478439331055,
0.0006308004958555102,
0.19232134521007538
] |
[
0.9954144954681396,
0.004031105898320675,
0.00045253447024151683,
0.00010181924153584987
] |
en
| 0.999997 |
Interestingly, although stiffness did not significantly affect acceleration and power density, there was a slight advantage of the stiffness of platform A matching the stiffness of the grasshopper, suggesting that there was an optimum platform stiffness for maximum acceleration and power density, when the mass of the platform is less than the 6.34 g platform tested here. Gilman et al. researched leaf compliance as part of their study into jumping by green anole lizards ( Anolis carolinensis ) and found that leaves had a stiffness ranging between 0.03–1.43 m N −1 , although the species of leaf was not recorded. The measurement of leaf stiffness, which involved the application of external forces, was dependent on factors such as osmotic pressure, which varied with age of the leaf and season .
|
39676724_p42
|
39676724
|
Connections between acceleration and power density
| 3.940441 |
biomedical
|
Study
|
[
0.9254104495048523,
0.0005560659919865429,
0.0740334689617157
] |
[
0.9988883137702942,
0.0006592096760869026,
0.00040245597483590245,
0.00005005611092201434
] |
en
| 0.999996 |
The mechanical structure of leaves is formed of multiple cantilevers, i.e. a mass extending horizontally that is supported at only one fixed end . Therefore, the interaction between the plant and an external mass (whether this be experimentally or animals jumping from it) becomes highly complex and difficult to measure. On a cellular level, the stiffness of leaves depends on the number of cells in a tissue sample, because deformation is resisted by the cell walls . Owing to the variation of leaf stiffness and the complex relationship between a leaf and a jumper, it is difficult to suggest whether this specific platform mass and stiffness that produced an optimum acceleration and power is applicable to grasshopper locomotion in the field.
|
39676724_p43
|
39676724
|
Connections between acceleration and power density
| 3.631033 |
biomedical
|
Study
|
[
0.943347156047821,
0.00033535499824211,
0.056317541748285294
] |
[
0.9633556604385376,
0.035379063338041306,
0.0011123210424557328,
0.0001529936125734821
] |
en
| 0.999994 |
It has been demonstrated that grasshoppers have the ability to recover energy from a recoiling compliant substrate, thus maintaining high velocity, acceleration and power. Biological LaMSA jumpers can successfully exploit their environment for jumping, although further exploration is necessary to conclude the mechanisms that underpin the ability of grasshoppers to do so. This research has broadened our knowledge around the incredibly energy efficient grasshopper jumping mechanisms and how they have evolved to adapt to uneven substrates.
|
39676724_p44
|
39676724
|
Conclusion
| 2.10005 |
other
|
Study
|
[
0.48526012897491455,
0.0011885844869539142,
0.5135512948036194
] |
[
0.5255833864212036,
0.45815587043762207,
0.01515184249728918,
0.0011088998289778829
] |
en
| 0.999998 |
In the past decade, cryo–electron microscopy (cryo-EM) single-particle analysis (SPA) has become a major technique for structural biology ( 1 ), excelling in areas where protein crystallography has been limited ( 2 ). The technical infrastructure needed for this endeavor involves specialized cryo-EM microscopes with a high-coherence electron field emission gun (FEG), high operational vacuum for the stability of the FEG, improved columns, constant-power lenses, and specialized cryo-stages. Together, these provide the optical and mechanical stability required for obtaining a typical high-resolution SPA dataset during 12- to 24-hour automated data collection. Given that improvements in electron source, column, and optics were already available for materials sciences, the development of direct electron detectors is likely the most critical innovation in aiding low-dose imaging and achieving the “resolution revolution” in cryo-EM ( 3 ). Pushing the signal-to-noise level beyond film, originally used in early cryo-EM studies, these detectors were in first instance designed to yield optimal detective quantum efficiency (DQE) for imaging 200- to 300-keV electrons ( 4 – 6 ). These detectors also made it possible to fractionate the total dose required for a typical exposure (50 to 60 e − Å −2 ) into subframes. This, combined with the advancements in the motion correction algorithm ( 7 ), facilitated the efficient correction of beam-induced sample drift, as well as residual stage drift in the individual exposures. This made it possible to routinely achieve resolutions below 4 Å for biological specimens ( 3 ). In addition, many notable innovations in recent years have boosted the achievable resolution in cryo-EM. This includes the incorporation of cold FEGs for cryo-EM ( 8 ), the development of a stable energy filter to enable zero loss filtering using a narrow slit width of 10 eV or below ( 9 ), and the improvements in minimizing detector readout noise, namely, correlated double sampling (CDS) mode in Gatan’s K3 detector ( 10 ) and multiframe CDS implementation in Falcon 3-4 detectors (Thermo Fisher Scientific). Individually or combined, these innovations have made it possible to solve biological structures to sub–2-Å resolution ( 11 – 15 ).
|
39752481_p0
|
39752481
|
INTRODUCTION
| 4.751628 |
biomedical
|
Review
|
[
0.994929850101471,
0.0021891985088586807,
0.002880895510315895
] |
[
0.15441285073757172,
0.00335480528883636,
0.8410751223564148,
0.0011572444345802069
] |
en
| 0.999997 |
For most single-particle projects, the total imaging time required for their successful completion can be divided into screening and optimizing sample preparation conditions and the final high-resolution imaging. Screening usually involves assessing grid preparation by imaging multiple grids to identify the conditions that result in optimal particle distribution and quality of ice. During the screening, a limited dataset from two to three grid squares with varying ice thickness is assessed for best particle distribution from each grid. This exercise is also useful to assess potential pitfalls in cryo-EM grid preparation such as preferential orientation and air-water interface or to understand problems arising from biochemistry like intactness of biological complexes and flexibility ( 16 – 18 ). Ideal cryo-EM samples would typically yield class averages showing multiple orientations as well as secondary structure level details at ~6- to 8-Å resolution. A three-dimensional (3D) reconstruction from such a dataset could potentially yield 3- to 6-Å resolution, enough to screen for complex formation and antibody or fab binding and even for the presence of peptides or small molecules (at ~3- to 4-Å resolution). Once the grid-making conditions have been established from limited processing, the same grid or replicates can be sent for imaging at state-of-the-art 300-keV microscopes such as the TFS Titan Krios or the Jeol Cryo Arm 300 for the collection of large highest-quality datasets ( 19 ).
|
39752481_p1
|
39752481
|
INTRODUCTION
| 4.218413 |
biomedical
|
Study
|
[
0.9994171857833862,
0.00024210552510339767,
0.00034078105818480253
] |
[
0.9921420812606812,
0.0032249907962977886,
0.004501591436564922,
0.00013132643653079867
] |
en
| 0.999996 |
The screening process can easily amount for the largest proportion of imaging time required for the successful completion of a project, especially for more challenging samples. It should be noted that higher-quality datasets for screening purposes require relatively expensive cryo–transmission electron microscopes (cryo-TEMs) typically having a 200-keV FEG source, constant-power lenses, direct electron detectors, and often autoloader stages. In this context, lately, potential benefits have been highlighted for performing cryo-EM at 100 keV ( 20 ) and the application for high-resolution structure determination was shown using a modified DECTRIS EIGER X 0.5-megapixel detector on a TF20 operated at 100 keV ( 21 ). Since then, a dedicated 100-keV microscope was developed, featuring a 1-megapixel dedicated 100-keV detector (DECTRIS SINGLA), a low chromatic aberration coefficient (C c ) objective lens, and a Schottky FEG (S-FEG) electron source with compact power supply. The design of this microscope, especially in its nonrequirement of the greenhouse gas sulfur hexafluoride (SF 6 ) for insulation, considerably reduced the cost, complexity, and environmental impact ( 22 ). This system is not yet commercially available, but these studies show the promising future of cheaper microscopy solutions enabling democratization of the field ( 16 , 22 , 23 ). Taking a lead from these studies, commercial manufacturers have made strides into developing sub–200-keV detectors. This includes the Alpine direct electron detector by Gatan, Falcon-C by Thermo Fisher Scientific, and Quantum C100 from Quantum Detectors. Recently, a microscope by Thermo Fisher Scientific called Tundra has been shown to be an effective tool for sample screening using 100-keV imaging with the Falcon-C detector ( 23 ). However, this imaging configuration is still a relatively expensive microscope because of its X-FEG source, higher-quality column with a constant-power objective lens, and semiautomated sample loader. One of the main bottlenecks for wide adoption of single-particle cryo-EM hence is the high cost of these imaging capabilities as well as the service contracts to maintain them.
|
39752481_p2
|
39752481
|
INTRODUCTION
| 4.374211 |
biomedical
|
Study
|
[
0.9985677003860474,
0.0003281820681877434,
0.0011041414691135287
] |
[
0.8599705100059509,
0.0028324988670647144,
0.1369026005268097,
0.0002944813168141991
] |
en
| 0.999997 |
In the context of commercially available sub–200-keV optimized direct electron detectors, we looked for further opportunities for making high-resolution cryo-EM more affordable. We focused especially on the sample screening aspect of cryo-EM SPA, which requires the ability to reach a resolution of 3 to 4 Å in the best-case scenario. However, so far, these detectors have only been tested on costlier FEG source instruments but not yet on standard 120-keV TEMs with a common lanthanum hexaboride (LaB 6 ) thermionic source. There are few published cases where LaB 6 filaments were used to obtain 6- to 10-Å–resolution structures imaged at 400 keV using a Jeol JEM-4000 ( 24 – 26 ), but these studies predate the introduction of direct electron detectors and were imaged on film. The major inherent factor limiting resolution is the spatial and temporal coherence of the LaB 6 source. The energy spread for the LaB 6 source is typically in the range of 1.5 to 2 eV as opposed to <0.8 to 0.9 eV for Schottky FEG ( 27 ), 0.7 eV for X-FEG, and <0.3 eV for cold FEG ( 9 ). However, even taking these limitations into account, the line and point resolution of typical 120-keV LaB 6 TEMs is in the range of 2.0 and 3.6 Å, respectively. In addition, practical limitations such as the substantial sample drift of side-entry cryo-transfer holders, insufficient environmental isolation, and stability in terms of vibration and acoustic noise can further limit the performance. Since dose-fractionated data from a direct electron detector can be used to mitigate problems with stage drift and beam-induced motion, we envisaged that the addition of such a detector on a traditional 120-keV LaB 6 microscope can hugely boost its performance. For the present study, a standard Tecnai 120-keV LaB 6 G2 Spirit TWIN microscope (Thermo Fisher Scientific) was retrofitted with Gatan’s sub–200-keV optimized direct electron detector Alpine. At 100 keV, the Alpine detector has been shown to have fourfold better DQE than Gatan’s K3 at the Nyquist frequency ( 28 ). We characterize this imaging configuration for its utility to derive high-resolution structural data. In addition to well-behaved samples, we show that challenging proteins as well as sub–100-kDa proteins can be resolved to higher resolution using this configuration. This could be a more affordable and widely available option for high-quality cryo-EM SPA for both screening purposes and structure determination.
|
39752481_p3
|
39752481
|
INTRODUCTION
| 4.346189 |
biomedical
|
Study
|
[
0.9991945624351501,
0.0004092561430297792,
0.00039618529262952507
] |
[
0.999189555644989,
0.00023548645549453795,
0.0004964285180903971,
0.00007850045949453488
] |
en
| 0.999996 |
To test the impact of a direct electron detector on the performance of a standard 120-keV cryo-capable TEM, we used a Tecnai G2 Spirit with a TWIN objective lens (non–constant-power lens) with a spherical aberration coefficient (C s ) of 2.2 mm and a Cc of 2.2 mm. The electron source used is a standard LaB 6 thermionic source with a 15-μm flat tip and a 90° cone angle (DENKA). As these instruments are not typically used for high-resolution data collection, the housing conditions for such a microscope are usually not optimal. With low-DQE charge-coupled device (CCD) cameras traditionally fitted with these microscopes combined with poor vibration isolation and poor stage drift, observing features close to a line resolution of 2.0 Å tends to be difficult but not impossible.
|
39752481_p4
|
39752481
|
120-keV LaB 6 TEM performance aided by a sub–200-keV direct electron detector
| 4.149457 |
biomedical
|
Study
|
[
0.9992818236351013,
0.00022636898211203516,
0.0004918146878480911
] |
[
0.9988252520561218,
0.0008736835443414748,
0.00023216204135678709,
0.00006890772056067362
] |
en
| 0.999996 |
After tuning the microscope for 120 keV, a cross-grating sample mounted on a room-temperature holder was imaged to assess the optical performance of the microscope under parallel illumination conditions. Once the stage drift settled, it was possible to observe a 2.35-Å lattice signal as observed in selected-area fast Fourier transform from a high-resolution transmission electron microscope image captured using a BM Eagle CCD detector . The 4-s exposure did result in capturing some residual drift as seen in a directional loss of information in the power spectrum . This camera was unmounted and was replaced with the commercially available Gatan sub–200-keV optimized Alpine detector . As it is a direct electron detector, imaging with the Alpine detector allowed us to save images as dose-fractionated movies, which allows for postacquisition drift correction. Optical alignments of the microscope were checked, and a longer exposure of 8 s was chosen to capture stage drift and environmental vibrations that might affect data collection during typical usage. The resultant exposure showed severe motion blurring due to drift , which is also visible in the associated power spectrum . The digital micrograph’s built-in motion correction tools were used to analyze and further correct the drift after acquisition. The drift analysis showed that there was ~1.6-nm cumulative drift in the y direction and ~0.2-nm drift in the x direction. The power spectrum from the motion-corrected aligned average clearly showed the gold ⟨111⟩ (2.35 Å) and ⟨200⟩ (2.04 Å) signals as shown in . The same tests were done with a Gatan 626 holder with and without liquid nitrogen (LN 2 ) to test for holder performance. In the first instance, we observed washout of Thon rings in the power spectrum caused by vibration (not fixable by drift correction). These vibrations, predominantly acoustic in nature, were seen to have more pronounced impact with cryoholders likely because of the larger surface area associated with LN 2 dewars compared to the room-temperature holder . The Alpine camera power supply was identified as the reason for the acoustic vibrations and was relocated to an adjacent utility room, thereby attenuating these vibrations and increasing data quality . Next, we proceeded to automate the usage of the Alpine camera using SerialEM ( 29 ) as well as to enable automatic data collection. We performed tests with and without beam image shift on a C-Flat grid (Protochips) to mimic single-particle acquisition for collecting nine holes per stage move. The purpose of this was to test whether switching between lower selected-area (SA) to high SA magnifications required for cycling between hole finding, autofocusing, and data acquisition would lead to any major hysteresis in the non–constant-power TWIN objective lens and/or requires longer beam settling time after image shift. The results of contrast transfer function (CTF) estimation using CTFFIND ( 30 ) showed that ~90% of the images could be fit to a resolution of 3.5 to 4 Å and had an average astigmatism measure of 29.2 nm . Coma-free alignment is critical for obtaining high-resolution cryo-EM datasets ( 31 ). In contrast to higher-end Thermo Fisher Scientific microscopes where coma-free alignment is accessible in the user interface, only rotation centering is available on the Tecnai G2 Spirit. This issue is easily mitigated by integrating the camera use with SerialEM ( 29 ), which makes it possible to access the tilt coils to achieve coma-free alignment as well as to perform coma versus image-shift calibration to reduce astigmatism and coma arising from using beam image shift. Once these calibrations were performed and sufficiently reproducible optical performance was observed, we proceeded to collect SPA datasets.
|
39752481_p5
|
39752481
|
120-keV LaB 6 TEM performance aided by a sub–200-keV direct electron detector
| 4.298289 |
biomedical
|
Study
|
[
0.9989425539970398,
0.00042998060234822333,
0.0006274429033510387
] |
[
0.9994033575057983,
0.0002440177631797269,
0.00028379212017171085,
0.00006879943248350173
] |
en
| 0.999996 |
To assess whether the optical stability seen in a limited number of datasets measured on carbon translates to long data collections required for biological specimen, we used apoferritin as our first test sample. We obtained a throughput of ~170 images/hour with beam image-shift data collection on nine holes and a 15-s stage drift settling time, which was used for four rounds of autofocusing. For the whole dataset, the maximum resolution to which the CTF could be reliably fit using CTFFIND was Gaussian distributed around 4.5 to 5 Å. Further image processing yielded a 2.65-Å final structure . Although the first 20,000 particles were sufficient to reach 2.92 Å, further addition of particles did not markedly improve the result. The calculated B -factor from the Rosenthal and Henderson ( 32 ) plot was 139 Å 2 , and the sharpening B -factor from the Guinier plot calculated during refinement was 117 Å 2 . Although, before data collection, the coma was corrected to a threshold of 0.1 mrad, the average residual coma measured from CTF refinement ( 33 ) for the nine beam image-shift optics group was noted to be 0.58 and 0.60 mrad in x and y , respectively. Furthermore, analysis of Bayesian polishing ( 34 ) B -factor showed a steep rise in the estimated B -factor as a function of dose. We compared this to polishing B -factor obtained for a megapixel-equalized zero loss–filtered dataset of apoferritin acquired on the S-FEG 300-keV Titan Krios captured using a Gatan K3 detector at a pixel size of 0.82 Å. At 120 keV, the low-resolution contrast for non–dose-weighted averages was seen to drop progressively with every accumulated dose of 10 e − Å −2 . The low-resolution contrast for the 300-keV dataset, albeit being poor when compared to corresponding dose fractions from the 120-keV dataset, followed the same pattern of attenuation with the progression of dose . However, for high-resolution information, the resultant reconstructions showed that in the 120-keV dataset compared to the 300-keV dataset, very little high-resolution information is present after 30 e − Å −2 . These observations are in line with previous studies comparing 100- and 300-keV dose-dependent loss of high-resolution signal in the diffraction pattern from two-dimensional (2D) crystals of C 44 H 90 paraffin and purple membrane ( 20 ).
|
39752481_p6
|
39752481
|
Sub–3-Å structure of apoferritin
| 4.37349 |
biomedical
|
Study
|
[
0.9992387294769287,
0.000457053683931008,
0.00030419029644690454
] |
[
0.9991528987884521,
0.0002174654946429655,
0.0005213557742536068,
0.00010824641503859311
] |
en
| 0.999999 |
We further tested the imaging configuration to see how it fares with a more challenging asymmetric sample and its feasibility for screening. For this purpose, we chose the 153-kDa M 4 muscarinic acetylcholine receptor (M 4 mAChR) bound to the G i1 protein and in complex with the small molecule iperoxo as well as LY298 (M 4 R-G i1 -Ipx-LY298), which belongs to the group of G protein–coupled receptors (GPCRs). This complex was previously reconstructed to a resolution of 2.4 Å ( 35 ) from ~6000 zero loss–filtered movies collected at 300 keV on a Titan Krios with a K3 detector in CDS mode. Using the Alpine on the Tecnai G2 Spirit, we collected a megapixel-equalized dataset to match a typical screening dataset that we normally collect on a Talos Arctica. In ~11 hours, we obtained ~1300 movies. Initial analysis using a combination of RELION-4.0 and CryoSPARC ab initio classification followed by nonuniform refinement yielded a resolution of 5.8 Å. Further 3D classification in RELION-5.0 with BLUSH regularization ( 36 ) followed by refinement using BLUSH regularization yielded a final total of 50,031 particles that reached a resolution of 4.4 Å . Local resolution estimates showed that the resolution in the transmembrane helices (TMs) is of the order of 4.2 to 5 Å with the highest resolution of the order of 4.2 to 4.8 Å for TM3. This resolution in the TMs was sufficient to unambiguously trace the backbone as shown by the rigid-body fit of the previously published M 4 R-G i1 -Ipx-LY298 structure . The resolution was sufficient to unambiguously identify a notable tilt in the receptor with respect to the G protein when compared to that solved previously . The structure had sufficient resolution to identify the presence of the small-molecule (LY298) occupancy in the allosteric binding site, whereas the orthosteric inhibitor iperoxo was not visible . Notably, in the previously published reconstruction from the 300-keV dataset ( 35 ), a density for iperoxo was detectable, but the alkyne bond was not visible. Since the 300-keV dataset had 617,793 particles compared to the 50,031 particles in the current study, it is difficult to ascertain whether the lack of the iperoxo density in the structure is due to a poorer signal-to-noise (S/N) ratio of individual particles compared to the zero loss–filtered 300-keV FEG dataset or is a result of partial occupancy or dynamics, which would have benefited from averaging more particles. Although the resolution is not sufficient to unambiguously confirm the identity of LY298, the results show that this resolution level is sufficient to identify the presence or absence of a small ligand at the binding site.
|
39752481_p7
|
39752481
|
High-resolution structure of a 153-kDa asymmetric dynamic membrane protein
| 4.391964 |
biomedical
|
Study
|
[
0.999235987663269,
0.0005032630870118737,
0.0002607539063319564
] |
[
0.9991645812988281,
0.0003094055864494294,
0.00038469713763333857,
0.00014134756929706782
] |
en
| 0.999998 |
We further characterized the setup for imaging protein with a molecular mass smaller than 100 kDa. For this purpose, we chose human hemoglobin, a ~64-kDa heterotetrametric heme-containing protein. The data collection strategy was kept similar to that described for the GPCR and apoferritin samples. Hemoglobin particle distribution was less than ideal with more particles present toward the lip of the holes. About 1300 movies were collected between two different grids. Movies with the maximum observable resolution as reported by CTFFIND that exceeded 6 Å were discarded. This resulted in ~700 usable movies. This dataset was sufficient to get 2D class averages with secondary structure–level detail. For the 3D refinement, the conventional ab initio classification with a 30-Å initial search resolution cutoff proved to be intractable to result in a sufficient-quality ab initio model. Instead, an initial low-resolution cutoff of 12 Å and a high-resolution limit of 4 Å proved to be necessary to give a robust ab initio classification result. Particles corresponding to the resultant ab initio class that had secondary structure detail were then further subjected to homogeneous refinement to get a ~6-Å–resolution structure. This structure was used as an initial model for 3D classification using BLUSH regularization in RELION-5.0. Multiple classification rounds were required to home in on a final set of ~8000 particles, which contributed to the final reconstruction with 4.33-Å resolution [gold standard Fourier shell correlation (FSC) (0.143)]. Local resolution estimates showed that outer helices in both the α and β chains were exhibiting the lowest resolution of ~5.3 Å, whereas the highest resolution of 4.0 Å was seen for the inner helices . The density of heme and coordination with His 87 were also clearly visible as seen by the rigid body–docked Protein Data Bank (PDB) ID: 5NI1 to the EM density .
|
39752481_p8
|
39752481
|
A 4.33-Å structure of a 64-kDa protein hemoglobin
| 4.424785 |
biomedical
|
Study
|
[
0.999277651309967,
0.0005235837888903916,
0.00019886788504663855
] |
[
0.998904824256897,
0.000331182440277189,
0.0006008668569847941,
0.0001631448685657233
] |
en
| 0.999997 |
Our experiments show that a ubiquitously available 120-keV LaB 6 electron microscope such as a Tecnai G2 Spirit TWIN can be retrofitted with a sub–200-keV optimized direct electron detector to massively boost its performance level. The cost associated with procuring as well as maintaining these standard 120-keV microscopes is much lower than a modern autoloader system like a Thermo Fisher Scientific Talos Arctica or Glacios or Jeol Cryo Arm 200 or even FEG-based side-entry holder–driven microscopes such as the Talos F200X or Jeol JEM-F200. We were able to successfully integrate the Alpine detector and make it amenable for automated data collection through SerialEM and achieve a working throughput of roughly 170 movies/hour in CDS mode. The 2.65-Å structure of apoferritin achieved using this setup shows that with proper sample preparation, the datasets have enough optical quality to reach a resolution where model building is possible. The results from the Bayesian polishing routine from RELION ( 34 ) showed that the polishing B -factor increases more steeply for the 120-keV dataset compared to the 300-keV dataset . We do note here that the 300-keV dataset was collected using zero loss filtering, which is known to increase S/N ratio across the resolution spectrum ( 37 ). The progressive reduction in high-resolution signal per frame in the 120-keV dataset is likely due to the beam damage caused by the lower-keV electrons combined with a steep CTF envelope owing to poorer coherence of the LaB 6 source. Since the lower-keV images have very good low-resolution contrast but very little high-resolution information available past 30 e − Å −2 , the typically used total dose of 50 to 60 e − Å −2 for 200- to 300-keV imaging may not be required for 120-keV data collection. The lower dose requirement would further shorten the exposure time required to 3 to 4 s, allowing for an increase in throughput. This would be highly beneficial for challenging asymmetric or flexible targets where increasing the number of particles available for characterization becomes important. These observations are consistent with previous studies ( 20 , 21 ) where it has been established that 100-keV electrons have more signal-producing elastic scattering as opposed to 300 keV for a given dose. Collecting more data with shorter exposure times at a lower dose is hence a better data collection strategy for imaging at 120 keV. Although the non-CDS mode imaging is not tested in the present study, once used, it should further improve throughput.
|
39752481_p9
|
39752481
|
DISCUSSION
| 4.331522 |
biomedical
|
Study
|
[
0.9992307424545288,
0.0004376689321361482,
0.0003316002548672259
] |
[
0.9991449117660522,
0.00027246869285590947,
0.0004862428759224713,
0.00009639906784286723
] |
en
| 0.999996 |
The ability of the present imaging configuration to get to a resolution of 4.4 Å for challenging samples like the 153-kDa GPCR (M 4 R-G i1 -Ipx-LY298) as well as its ability to successfully achieve a 4.33-Å reconstruction for an even smaller 64-kDa protein target like the C2 symmetric human hemoglobin complex, with sparse data, shows the potential of this configuration to achieve below 3- to 4-Å resolution by collecting larger datasets. The resolution achieved in both cases was enough to ascertain the presence or absence of ligands within the complex, especially as seen with the heme complex in the hemoglobin molecule obtained with only 8865 particles . The latest developments in the motion correction algorithm in MotionCor3 as well as RELION-5’s BLUSH regularization technique were critical for 3D classification to glean more information from these sparse datasets. Furthermore, the contrast available at low defocus makes imaging at lower keV an ideal candidate for imaging low–molecular weight protein targets as shown in fig. S4. There is sufficient low-resolution contrast to visually identify particles even at defocus as low as ~0.6 μm for GPCR as well as hemoglobin. The above results show that this imaging configuration is an effective SPA screening tool and, if needed, with larger datasets, can be used to get to a resolution of 3 Å where one can confidently do model building. Samples such as GPCRs, which require screening and characterizing not only on grid square–level variations but also on hole-to-hole–level variations, greatly benefit from screening on TEMs with a multisample loading system. Development of multigrid side-entry cryo-capable holders, which can accommodate three to four grids, would be game changing for using existing 120-keV microscopes for screening purposes. Automated refilling or longer holding times for cryo-conditions to be maintained overnight without the need for manual refill would also allow for longer data collection.
|
39752481_p10
|
39752481
|
DISCUSSION
| 4.382711 |
biomedical
|
Study
|
[
0.9994695782661438,
0.0002866007562261075,
0.00024394736101385206
] |
[
0.9975482821464539,
0.0005695209256373346,
0.0017776573076844215,
0.00010448094690218568
] |
en
| 0.999997 |
Widely available low-cost 120-keV LaB 6 microscopes once enabled with an affordable sub–200-keV direct electron detector like the Gatan Alpine detector can massively bring down the economic entry barrier of cryo-EM. This could provide a pathway to cryo-EM, particularly for institutions and countries that currently have no cryo-EM capability. At the same time, dedicated central cryo-EM facilities would receive more optimized samples, allowing for more efficient usage of high-end instrument time. In summary, the upgrade of accessible 120-keV LaB 6 TEMs could truly democratize cryo-EM with the biggest impact likely on the widespread accessibility of sample screening and optimization.
|
39752481_p11
|
39752481
|
DISCUSSION
| 3.647374 |
biomedical
|
Other
|
[
0.994186282157898,
0.0006047958158887923,
0.00520891509950161
] |
[
0.055212948471307755,
0.9380578994750977,
0.006300225388258696,
0.0004289614735171199
] |
en
| 0.999996 |
The Alpine detector was bottom mounted as shown in fig. S1 (A and B). The camera was allowed to access the Tecnai G2 Spirit TWIN prespecimen blanker (gun deflector) via a direct connection from TEM lens control electronics to the Gatan camera controller. This allowed the camera to interact with the microscope as a standalone camera. Further integration of the camera with the microscope was performed through SerialEM ( 29 ) (ver. 4.0.28). Standard SerialEM calibration was carried out to efficiently use the camera along with the microscope.
|
39752481_p12
|
39752481
|
MATERIALS AND METHODS
| 3.720872 |
biomedical
|
Study
|
[
0.9974271655082703,
0.0003609073755796999,
0.002211980288848281
] |
[
0.8587260246276855,
0.14010365307331085,
0.0007602022960782051,
0.00041016528848558664
] |
en
| 0.999996 |
A Tecnai G2 Spirit TWIN (Thermo Fisher Scientific) equipped with standard LaB 6 thermionic emission sources (DENKA LaB 6 , 15-μm-radius microflat surface with a 90° cone angle) was operated at 120 keV. A filament heat setting of 27 and a bias setting of 1 were chosen, which resulted in a gun emission of ~4.8 μA. A C2 aperture of 50 μm along with a spot size of 7 was chosen, and the microscope was operated in nanoprobe mode. Diffraction grating replica was used for pixel size calibration in a Gatan digital micrograph. A nominal magnification of 60,700× at camera level corresponding to a calibrated pixel size of 0.4055 Å (super-resolution mode) was chosen for data collection, and the beam was set to satisfy parallel illumination (C2 lens strength ~32.4%). This resulted in a beam with a diameter of ~1.5 μm. The camera was operated in CDS mode. All resultant movies were Fourier binned (2×), dose weighted, and motion corrected using UCSF Motioncor2 (version 1.6.3) ( 7 ) to output both dose-weighted and non–dose-weighted averages with a resultant pixel size of 0.811 Å. An amplitude contrast of 0.1 was used. The non–dose-weighted averages were used for estimating CTF parameters using CTFFIND 4.1.8 ( 30 ), using a RELION-4.0.1 ( 38 ) wrapper.
|
39752481_p13
|
39752481
|
Data collection
| 4.279401 |
biomedical
|
Study
|
[
0.9991657733917236,
0.00042259730980731547,
0.00041168698226101696
] |
[
0.9979914426803589,
0.0014460455859079957,
0.0004155395436100662,
0.00014696196012664586
] |
en
| 0.999998 |
Movies were recorded at a dose rate of ~6.5 e − pixel −1 s −1 . A total exposure time of 7 s was used to accumulate a typical SPA dose of 60.4 e − Å −2 . Automatic data collection using SerialEM was performed to collect a series of test images. A low-magnification map was collected on a relatively flat region of a Au cross-grating (Ted Pella Inc.). Points for image acquisition were set 1 to 2 μm apart for automated data collection to mimic the SPA style data collection routine. A drift settling time of 15 s spent doing autofocusing was used to minimize residual stage drift before data collection. On-axis data collection and beam image-shift data collection using nine holes acquisition per stage shift pattern with and without coma versus image-shift compensation were performed to test for instabilities in imaging that might arise from optics, environmental, or holder instabilities. Data were collected on the same Au cross-grating grid mounted on a room-temperature holder and a Gatan 626 holder at room temperature as well as at −172°C. Before high-resolution data collection, the LN 2 level was ensured to be below the temperature transfer rod in the cryoholder. Periodic LN 2 filling of the cryoholder was performed as required, ensuring the above conditions to be met.
|
39752481_p14
|
39752481
|
Holder and optics stability measurements
| 4.194268 |
biomedical
|
Study
|
[
0.9992140531539917,
0.0003598626935854554,
0.00042611482786014676
] |
[
0.9993041753768921,
0.0004369018424768001,
0.0001978072978090495,
0.00006113611743785441
] |
en
| 0.999997 |
Before imaging all cryo-EM samples, the astigmatism and coma were corrected and coma versus image-shift calibrations were done using SerialEM on a C-Flat grid (Protochips) mounted on a room-temperature holder. Once the beam settings were transferred to SerialEM for acquisition and autofocus setting, the room-temperature holder was removed and the cryo-EM grid was mounted on the stage using a Gatan 626 holder. The filament was kept running during the process.
|
39752481_p15
|
39752481
|
Single-particle data collection
| 4.013062 |
biomedical
|
Study
|
[
0.998143196105957,
0.0007829828537069261,
0.0010739101562649012
] |
[
0.6183047890663147,
0.37892118096351624,
0.0019189356826245785,
0.0008551423670724034
] |
en
| 0.999997 |
Three microliters of purified apoferritin from Thermo Fisher Scientific (VitroEase Apoferritin Standard) at a concentration of 4 mg/ml was applied on UltraAuFoil ( 39 ) R 1.2/1.3 300 mesh. The grids were glow discharged for 30 s at 30 mA in an atmosphere before application of the sample. The excess protein solution was blotted off using a blot force of −1 and a blot time of 3 s in a Mark IV Vitrobot (Thermo Fisher Scientific) at 100% humidity and 4°C.
|
39752481_p16
|
39752481
|
Grid preparation
| 4.091306 |
biomedical
|
Study
|
[
0.9993317723274231,
0.0003842469013761729,
0.00028406071942299604
] |
[
0.9766936898231506,
0.022200366482138634,
0.0007482761866413057,
0.0003576839226298034
] |
en
| 0.999996 |
Beam image-shift data collection with nine holes acquisition per stage shift was performed with coma compensation on. Movies were collected with a total dose of 50.3 e − Å −2 accumulated over 6.5-s exposure time at a dose rate of 5.25 e − pixel −1 s −1 fractionated into 52 frames. A total of 923 movies were collected in a time frame of 5.45 hours.
|
39752481_p17
|
39752481
|
Data collection
| 3.999283 |
biomedical
|
Study
|
[
0.9975321292877197,
0.0003491827810648829,
0.0021187046077102423
] |
[
0.9967857599258423,
0.002942320192232728,
0.00018588270177133381,
0.00008606853953097016
] |
en
| 0.999998 |
Particle picking was performed using Gautomatch (0.53) ( https://github.com/JackZhang-Lab/Gautmatch/ ) on the first 100 images and extracted and binned four times using RELION-4.0.1. These particles were then subjected to 2D classification followed by ab initio and 3D refinement in CryoSPARC 4.2.0 ( 40 ). The coordinates were then exported back to RELION-4.0.1 using pyem v0.5 and re-extracted in RELION-4.0.1 centered on refined coordinates. The particles were then used for training using Topaz picker ( 41 ), and the trained model was used for particle picking on the full dataset. The resultant particles were then extracted and binned four times and were imported to CryoSPARC 4.2.0 ( 40 ). After 2D classification and 3D homogeneous refinement, the final set of particles (127,118) was arrived upon ( Table 1 ). This final set of particles was then extracted similarly as described above at native pixel size and was further processed in CryoSPARC 4.2.0 ( 40 ). The refined particles were then subjected to Bayesian polishing in RELION-4.0.1 and were then further processed in CryoSPARC 4.2.0 ( 40 ). Several rounds of heterogeneous refinement to exclude noisy particles and CTF refinement yielded a 2.65-Å map (gold standard FSC 0.143 criteria). The pixel size calibration was verified compared to PDB ID: 7A6A.
|
39752481_p18
|
39752481
|
Image processing
| 4.372391 |
biomedical
|
Study
|
[
0.9993046522140503,
0.00046481643221341074,
0.0002305604430148378
] |
[
0.998672366142273,
0.0005900418618693948,
0.0005704169161617756,
0.00016718663391657174
] |
en
| 0.999997 |
Apoferritin was frozen as mentioned above, and data were collected using a G1 Krios (Thermo Fisher Scientific) operated at 300 keV, at a C2 aperture of 50 μm, and in energy-filtered TEM mode with imaging done on a Gatan K3 direct electron detector equipped with a Gatan BioQuantum energy filter. Imaging was performed at a nominal magnification of 105,000×, with zero loss filtering done using a 10-eV slit width. The K3 was operated in CDS mode with an effective pixel size of 0.82 Å. A dose rate of 8.9 e − pixel −1 s −1 was used to accumulate a total dose of 60.21 e − Å −2 over 5.0-s exposure time; this dose was further fractionated into 60 frames. Automatic data collection using beam image shift was performed using EPU software (Thermo Fisher Scientific).
|
39752481_p19
|
39752481
|
300-keV data collection
| 4.172155 |
biomedical
|
Study
|
[
0.9995050430297852,
0.0002729219268076122,
0.00022198091028258204
] |
[
0.9987345337867737,
0.0008691537659615278,
0.0003005416947416961,
0.00009581890481058508
] |
en
| 0.999997 |
Processing was done as mentioned above with the only exception that further refinement using shiny particles from the Bayesian polishing step was not performed. A final particle set of 106,568 particles resulted in 1.89 Å (gold standard FSC 0.143 criteria).
|
39752481_p20
|
39752481
|
300-keV data collection
| 3.537356 |
biomedical
|
Study
|
[
0.9979957342147827,
0.0005351384752430022,
0.0014691348187625408
] |
[
0.9703071713447571,
0.028877422213554382,
0.0004524850519374013,
0.000362884602509439
] |
en
| 0.999997 |
Both the 120- and 300-keV final sets of particles were re-extracted from non–dose-weighted motion-corrected averages corresponding to every 10 e − Å −2 increment. The dose-limited re-extracted particles were autorefined with their respective full dose final reconstructions as the initial model. Initial low-pass filtering for refinement was restricted to 12 Å. The estimated resolution from each dose-limited particle set along with its corresponding FSC (gold standard 0.143 criteria) plot is shown in Fig. 4B .
|
39752481_p21
|
39752481
|
Image processing
| 4.15543 |
biomedical
|
Study
|
[
0.9993615746498108,
0.00030419984250329435,
0.0003342715499456972
] |
[
0.9994754195213318,
0.00022487914247903973,
0.00024406374723184854,
0.00005568777851294726
] |
en
| 0.999999 |
Three microliters of M 4 R-G i1 -Ipx-LY298 purified as described previously ( 35 ) was applied on UltraAuFoil R 1.2/1.3 300 mesh at a concentration of 15 mg/ml. The grids were glow discharged for 180 s at 15 mA in an atmosphere before application of the sample. The excess protein solution was blotted off using a blot force of 10 and a blot time of 2.5 s in a Mark IV Vitrobot (Thermo Fisher Scientific) at 100% humidity and 4°C.
|
39752481_p22
|
39752481
|
Grid preparation
| 4.090124 |
biomedical
|
Study
|
[
0.9993771910667419,
0.00032627207110635936,
0.0002965456515084952
] |
[
0.9916359186172485,
0.007691515143960714,
0.0004712790250778198,
0.00020124575530644506
] |
en
| 0.999996 |
Beam image-shift data collection with nine holes acquisition per stage shift was performed with coma versus image shift–calibrated beam tilt compensation on. Data were collected in CDS mode with a super-resolution pixel size of 0.4055 Å. Movies were collected with a total dose of 60.53 e − Å −2 accumulated over 6.09-s exposure time at a dose rate of 6.74 e − pixel −1 s −1 fractionated into 52 frames. A total of 1324 movies were collected in a time frame of ~11 hours.
|
39752481_p23
|
39752481
|
Imaging
| 4.092816 |
biomedical
|
Study
|
[
0.9981226325035095,
0.0003403204900678247,
0.0015370147302746773
] |
[
0.9985576272010803,
0.001200626022182405,
0.0001778520963853225,
0.00006387531902873889
] |
en
| 0.999997 |
The resultant movies were Fourier binned (2×), dose weighted, and motion corrected using UCSF Motioncor3 ( 7 ) to output both dose-weighted and non–dose-weighted averages with a resultant pixel size of 0.811 Å. Particle picking was performed using Gautomatch (0.53) on the first 100 images and extracted and binned four times. These particles were then subjected to 2D classification followed by ab initio and 3D refinement in CryoSPARC 4.2.0. The coordinates were then exported back to RELION using pyem v0.5 ( 42 ) and re-extracted in RELION-5.0 centered on refined coordinates. The particles were then used for training using Topaz picker ( 41 ) through a RELION-5.0 wrapper, and the trained model was used for particle picking on the full dataset. This resulted in 479,634 particles, which were then extracted and binned four times and were imported to CryoSPARC 4.2.0. After 2D classification and 3D homogeneous refinement, the final set of 75,816 particles was arrived upon ( Table 1 ). This final set of particles was then re-extracted similarly as described above but with binning two times and was further processed in CryoSPARC 4.2.0. The refined particles were then subjected to Bayesian polishing in RELION-5.0 and were then reimported and further processed in CryoSPARC 4.2.0. Several rounds of heterogeneous refinement to exclude noisy particles and nonuniform refinement ( 43 ) yielded a 5.36-Å map (gold standard FSC 0.143 criteria).
|
39752481_p24
|
39752481
|
Image processing
| 4.429306 |
biomedical
|
Study
|
[
0.9990707635879517,
0.0005815416807308793,
0.00034769647754728794
] |
[
0.9985260963439941,
0.0006567672244273126,
0.00062700011767447,
0.00019021568004973233
] |
en
| 0.999996 |
The volume erase tool in UCSF Chimera ( 44 ) was used to remove the micelle density from the map, and a mask containing only the TM region and the G protein was prepared from this map in RELION-5.0 and was used for further 3D classification with BLUSH regularization on the full dataset. The highest-resolution class corresponding to 50,089 particles was then re-extracted at native pixel size and subjected to mask refinement with BLUSH regularization. This pushed the resolution to 4.8 Å (gold standard FSC 0.143 criteria). Polishing followed by per-particle defocus refinement further improved the map to achieve 4.38-Å resolution (gold standard FSC 0.143 criteria) .
|
39752481_p25
|
39752481
|
Image processing
| 4.301676 |
biomedical
|
Study
|
[
0.9994309544563293,
0.0003576091839931905,
0.00021138708689250052
] |
[
0.9979932308197021,
0.0012726287823170424,
0.000565210182685405,
0.00016890134429559112
] |
en
| 0.999997 |
Hemoglobin was purified from human blood collected from a healthy adult volunteer. Before the collection of blood, informed consent of the volunteer was obtained in accordance with 2022-30658-70864 approved by the Monash University Human Research Ethics Committee. Erythrocytes were isolated via centrifugation, and cell pellets were diluted in lysis buffer (50 mM tris and 200 mM NaCl, pH 7.4). To isolate hemoglobin, erythrocytes were lysed with a tight-fit dounce homogenizer and clarified by centrifugation at 18,000 rpm. Cell lysate was further purified using a Superdex 26/600 S200 size exclusion column (Cytiva) pre-equilibrated in size exclusion chromatography buffer (50 mM tris and 200 mM NaCl, pH 7.4). Fractions were collected, pooled, and concentrated to ~10 mg/ml using a 30-kDa spin filter column (Millipore) before being stored at −80°C.
|
39752481_p26
|
39752481
|
Protein purification
| 4.148604 |
biomedical
|
Study
|
[
0.999552309513092,
0.0002973586379084736,
0.0001502713857917115
] |
[
0.997745931148529,
0.001633227220736444,
0.0004584972048178315,
0.00016226738807745278
] |
en
| 0.999996 |
Three microliters of purified hemoglobin sample at a concentration of 10 mg/ml was applied on UltraAuFoil R 1.2/1.3 300 mesh. The grids were glow discharged for 30 s at 30 mA in atmosphere before application of the sample. The excess protein solution was blotted off using a blot force of −1 and a blot time of 3 s in a Mark IV Vitrobot (Thermo Fisher Scientific) at 100% humidity and 4°C.
|
39752481_p27
|
39752481
|
Grid preparation
| 4.073343 |
biomedical
|
Study
|
[
0.9991773962974548,
0.0005390128935687244,
0.00028354671667329967
] |
[
0.9543535113334656,
0.04404595121741295,
0.0010430044494569302,
0.0005575516843236983
] |
en
| 0.999998 |
Beam image-shift data collection with nine holes acquisition per stage shift was performed with coma versus image shift–calibrated beam tilt compensation enabled. Data were collected in CDS mode with a super-resolution pixel size of 0.4055 Å. Movies were collected with a total dose of 50.14 e − Å −2 accumulated over 6.711-s exposure time at a dose rate of 5.07 e − pixel −1 s −1 fractionated into 55 frames. A total of 1368 movies between two grids were collected.
|
39752481_p28
|
39752481
|
Imaging
| 4.116457 |
biomedical
|
Study
|
[
0.9986879229545593,
0.00038507013232447207,
0.0009270132286474109
] |
[
0.9988415837287903,
0.0009204453090205789,
0.00017017783829942346,
0.0000678113428875804
] |
en
| 0.999996 |
The resultant movies were Fourier binned (2×), dose weighted, and motion corrected using UCSF Motioncor3 ( 7 ) to output both dose-weighted and non–dose-weighted average with a resultant pixel size of 0.811 Å. Particle picking was performed using Gautomatch (0.53) on the first 100 images and extracted and binned four times. These particles were then subjected to 2D classification followed by ab initio and 3D refinement in CryoSPARC 4.2.0. The coordinates were then exported back to RELION using pyem v0.5 and re-extracted in RELION-5.0 centered on refined coordinates. The particles were then used for training using Topaz picker through RELION-5.0, and the trained model was used for particle picking on the full dataset. This resulted in 382,710 particles. In CryoSPARC 4.2.0, ab initio classification was performed using an initial low-resolution search cutoff of 12 Å, a high-resolution limit of 4 Å, and three classes with class similarity set to 0. Particles corresponding to the resultant ab initio class that had secondary structure detail were then further subjected to homogeneous refinement to get a ~6-Å–resolution structure. This structure was used as an initial model for 3D classification using BLUSH regularization in RELION-5.0. Multiple classification rounds were done to reach a final set of 8865 particles ( Table 1 ), which contributed to the final reconstruction with 4.33-Å resolution (gold standard FSC 0.143) . Visualization and rigid-body docking for all structures were performed using UCSF Chimera ( 44 ) (ver. 1.16).
|
39752481_p29
|
39752481
|
Image processing
| 4.40608 |
biomedical
|
Study
|
[
0.9991471767425537,
0.0005518721882253885,
0.00030093779787421227
] |
[
0.9984379410743713,
0.0006856707041151822,
0.0006890390650369227,
0.0001874225417850539
] |
en
| 0.999996 |
Studies comparing the mental health of students from different countries during the COVID-19 pandemic have found differences in anxiety and suicidal thoughts, 5 depressive and anxiety symptoms, 6 functional difficulties, stress, and concerns related to COVID-19 pandemic; 7 studies have also reported stress, anxiety and depression 8 in university students worldwide, indicating that students as a group are vulnerable to development of mental disorders and aggravation of pre-existing ones. For instance, a comparison between Portuguese and Brazilian university students showed higher levels of depressive symptoms among the latter than among the former. 9 Compared with university students from Turkey, Poland, Slovenia, Czech Republic, Ukraine, Russia, Israel and Colombia, German students presented a lower prevalence of anxiety symptoms. 6
|
39494847_p0
|
39494847
|
Cross-country studies
| 3.914338 |
biomedical
|
Study
|
[
0.9980974793434143,
0.0003506588109303266,
0.00155183847527951
] |
[
0.8001066446304321,
0.0015078485012054443,
0.1981726735830307,
0.0002127830230165273
] |
en
| 0.999997 |
Governments’ responses to COVID-19 have had a significant impact on the prevalence of depressive symptoms, with countries where governments implemented stringent policies promptly (according to the Oxford COVID-19 Government Response Index) showing lower depressive symptom prevalence. 10 This indicates that a rapid public health response, as well as reducing mortality rates, may protect mental well-being and prevent greater psychiatric morbidity by providing coping tools and resilience against uncertainty. 10 The Oxford COVID-19 Government Response Index is based on 13 metrics, including school closures and workplace closures, restrictions on public gatherings, closures of public transport, stay-at-home requirements, testing policy, face coverings and vaccine policy. The values of this composite measure range from 0 to 100 (100 = strictest). Brazil had a lower index (47.98 to 36.60) than Germany (49.14 to 39.12) throughout the data collection period of this study, except for the period from 11 December 2021 to 31 December 2021. 11 As there have been few cross-country studies 5 – 9 comparing university students’ mental health status in light of their governments’ responses to the COVID-19 pandemic and the changes in study and living conditions that occurred during the pandemic, further research is needed.
|
39494847_p1
|
39494847
|
Cross-country studies
| 3.922915 |
biomedical
|
Study
|
[
0.9966289401054382,
0.00024429953191429377,
0.0031268191523849964
] |
[
0.9962142109870911,
0.0006697140634059906,
0.0030613651033490896,
0.00005467091978061944
] |
en
| 0.999995 |
Brazil and Germany had different governmental responses to the COVID-19 crisis, especially regarding social distancing measures against COVID-19. Germany adopted a more stringent approach, involving stricter restrictions on public gatherings, internal and international travel controls, stay-at-home mandates, face covering requirements and testing policy. 12 By contrast, Brazil primarily relied on vaccination as its foremost national strategy to mitigate the transmission of COVID-19. 13 , 14 Therefore, this study aimed to investigate potential differences between university students from Brazil and from Germany with respect to (a) depressive symptoms and hazardous alcohol use, (b) social and emotional aspects (loneliness, self-efficacy, perceived stress, social support and resilience) and (c) attitudes towards vaccination. The choice of variables was guided by a previous study conducted by our research group, 12 which identified variations in these variables related to the period of the COVID-19 pandemic. 15 , 16 Based on previous results by our research group, 16 , 17 we hypothesised that Brazilian university students would show higher levels of depressive symptoms and more difficulties with social and emotional aspects, such as perceived stress, loneliness, social support and resilience. Differences between the Brazilian and German university students’ attitudes towards vaccination were analysed exploratively.
|
39494847_p2
|
39494847
|
Aims of the study
| 3.895011 |
biomedical
|
Study
|
[
0.9809296131134033,
0.0006529854726977646,
0.018417416140437126
] |
[
0.9996391534805298,
0.0001967810094356537,
0.00012607057578861713,
0.00003793116411543451
] |
en
| 0.999998 |
Data used in this study were obtained from two cross-sectional studies comprising a sample of Brazilian and German university students. Concerning the sample of Brazilian university students, a cross-sectional anonymous online survey was conducted with students at the Federal University of Parana (Brazil) between November 2021 and March 2022 (for more information, see Prado et al 17 ). The inclusion criteria were current enrolment as a university student and being 18 years old or older; no exclusion criteria were applied. The ethics committee of the Federal University of Parana granted approval for this study . The sample comprised N = 2437 participants. The Brazilian survey was conducted when no strict social distancing measures were in place in Parana state, and face-to-face academic activities in the university had resumed ; although academic and administrative activities were being developed remotely throughout the pandemic period, hybrid activities were allowed at the discretion of each department from September 2021. Vaccination against COVID-19 and wearing a face mask were mandatory to access university buildings. From the beginning to the end of the data collection, the full vaccination of the population in Brazil varied from 54.68 to 74.35%, and the reproduction rate of COVID-19 varied from 0.96 to 0.82. 17 The Federal University of Parana has around 39 000 students.
|
39494847_p3
|
39494847
|
Participants and procedures
| 3.526002 |
biomedical
|
Study
|
[
0.9708421230316162,
0.0005926406010985374,
0.02856525219976902
] |
[
0.999395489692688,
0.0004260186688043177,
0.000134886140585877,
0.00004361187166068703
] |
en
| 0.999996 |
Regarding the sample of German university students, a cross-sectional anonymous online survey was conducted with students of six universities (four of which were Universities of Applied Sciences – ‘ Fachhochschule ’ in German), in Saxony, Germany, between April and May 2022 (for more information, see Kohls et al 16 ). The inclusion criteria were current enrolment as a university student and being 18 years old or older; again, no exclusion criteria were applied. The ethics committee of the Medical Faculty of Leipzig University granted approval for the current study . The sample comprised N = 5474 participants. The German survey was conducted when no strict social distancing measures were in place, and face-to-face academic activities had resumed, with face mask mandates applied to university buildings. 11 From the beginning to the end of the data collection, the full vaccination status of the population in Germany ranged from 73.11 to 75.97%, and the reproduction rate of COVID-19 ranged from 1.20 to 0.81. 11 Altogether, the German universities had around 69 981 students.
|
39494847_p4
|
39494847
|
Participants and procedures
| 3.090064 |
biomedical
|
Study
|
[
0.979097843170166,
0.0007772716926410794,
0.02012486942112446
] |
[
0.9992731213569641,
0.000538534251973033,
0.0001351946557406336,
0.00005313322981237434
] |
en
| 0.999996 |
In both countries, participants were recruited using the official email and social media channels of the universities. The survey was set up in the online tool EFS Survey Unipark (version 21.1), a reliable assessment tool which adhered to the data protection guidelines in both countries. All participants provided online informed consent before participating via an opt-in function, where they were informed about the voluntary nature of participation and the guarantee of anonymity.
|
39494847_p5
|
39494847
|
Participants and procedures
| 1.750376 |
biomedical
|
Study
|
[
0.6524472832679749,
0.0017309922259300947,
0.34582170844078064
] |
[
0.802151620388031,
0.19619235396385193,
0.0009545495850034058,
0.0007015622686594725
] |
en
| 0.999997 |
This study draws on data and results from two congruent cross-sectional surveys conducted in Brazil 10 and Germany 11 that used identical questionnaires. The ENRICHD Social Support Instrument (ESSI) and the question on attitude towards vaccination were translated to Portuguese using a back-translation method. Further information on the psychometric properties of the instruments (including Cronbach's alpha) can be found in our previous publications. 16 , 17
|
39494847_p6
|
39494847
|
Measures
| 2.595139 |
biomedical
|
Study
|
[
0.9643428325653076,
0.001184687134809792,
0.03447256237268448
] |
[
0.9992044568061829,
0.00049530592514202,
0.00022941702627576888,
0.00007083149102982134
] |
en
| 0.999997 |
In both surveys, the following sociodemographic data were assessed: gender (female, male, diverse), age, relationship status (in a relationship or single), being a parent, residential status (living alone or with other people), study programme (bachelor, master or other) and migration background (self, parents or no migration background). In addition, chronic somatic diseases (‘Do you suffer from a (chronic) physical illness?’) and diagnosed mental disorders were assessed with ‘yes’ or ‘no’ answer options. Chronic somatic diseases were assessed because people with such conditions were at greater risk of dying from COVID-19 or suffering from long-term effects of COVID-19. COVID-19 vaccination status (‘What is your current COVID-19 vaccination status?’) was also assessed. Given the differences in the studies conducted in Germany and Brazil and the respective evolving COVID-19 vaccination scenarios, the answer options differed. For the Brazilian sample, they were ‘I am partly vaccinated (one shot)’, ‘I am fully vaccinated (two shots or single shot)’ and ‘I am not vaccinated’. This was because the Federal University of Parana required students to be fully vaccinated to be allowed to be present in face-to-face classes when they resumed; this did not occur in the German universities where the study took place. For the German sample, the answer options encompassed ‘I am fully vaccinated’, ‘I am not fully vaccinated, but would like to be vaccinated’ and ‘I am not fully vaccinated and do not want to be vaccinated’. The data were dichotomised into ‘fully vaccinated’ and ‘not vaccinated’, and Brazilian participants who reported being partly vaccinated ( n = 40, 1.6%) were excluded from the dichotomisation.
|
39494847_p7
|
39494847
|
Sociodemographic information, physical conditions and vaccination status
| 3.859265 |
biomedical
|
Study
|
[
0.9971931576728821,
0.0005920646945014596,
0.002214796142652631
] |
[
0.9995854496955872,
0.0002273465652251616,
0.00014694030687678605,
0.000040271708712680265
] |
en
| 0.999999 |
Depressive symptoms over the past 14 days were assessed using the Patient Health Questionnaire-9 (PHQ-9). 18 , 19 The PHQ-9 is a validated and widely used instrument for assessment of depression severity and comprises nine items on a four-point Likert scale from 0 = ‘not at all’ to 3 = ‘nearly every day’. 18 The sum score ranges from 0 to 27, with higher scores indicating higher levels of depressive symptoms. A sum score of 10 or more indicates clinically relevant depressive symptoms (‘moderate to severe’). 18 Item 9 (‘thoughts that you would be better off dead, or of hurting yourself’) indicates suicidal thoughts when answered with a score of ≥1 (i.e. reporting suicidal thoughts on several days or more during the past 14 days).
|
39494847_p8
|
39494847
|
Depressive symptoms
| 4.074216 |
biomedical
|
Study
|
[
0.9988468885421753,
0.0008989193011075258,
0.0002542307775001973
] |
[
0.9972061514854431,
0.001671169768087566,
0.000993181485682726,
0.00012943753972649574
] |
en
| 0.999997 |
Levels of alcohol consumption were assessed with the hazardous use subscale of the Alcohol Use Disorders Identification Test (AUDIT-C). 20 With a five-point Likert scale ranging from 0 = ‘never’ to 4 = ‘four or more times a week,’ the subscale assesses the frequency of drinking alcohol (‘How often do you have a drink containing alcohol?’), the typical quantity of alcohol consumed (0 = ‘one or two’ up to 4 = ‘ten or more’) and the frequency of consuming large quantities of alcohol (i.e. six or more drinks on one occasion; 0 = ‘never’ to 4 = ‘once a week’). The total hazardous use subscale sum score ranges from 0 to 12, with higher scores indicating higher alcohol consumption and related risk.
|
39494847_p9
|
39494847
|
Alcohol and drug consumption
| 3.920225 |
biomedical
|
Study
|
[
0.998688280582428,
0.00036853013443760574,
0.0009430871577933431
] |
[
0.9977957010269165,
0.0016719248378649354,
0.0004715514078270644,
0.000060933089116588235
] |
en
| 0.999995 |
One item of the AUDIT-C was rephrased to ‘drug or substance use’ (0 = ‘never’ to 4 = ‘four or more times a week’) and was used to assess the frequency of drug consumption.
|
39494847_p10
|
39494847
|
Alcohol and drug consumption
| 1.984411 |
biomedical
|
Study
|
[
0.9756105542182922,
0.0027572736144065857,
0.021632106974720955
] |
[
0.9539379477500916,
0.04471011087298393,
0.0007493198500014842,
0.0006026859628036618
] |
en
| 0.999997 |
Experience of loneliness was assessed using the UCLA three-item loneliness scale. 21 , 22 Each item was answered on a four-point Likert scale ranging from 0 = ‘never’ to 3 = ‘often’, with a total sum score ranging from 0 to 9. Higher scores indicate more experience of loneliness.
|
39494847_p11
|
39494847
|
Social and emotional aspects
| 2.640857 |
biomedical
|
Study
|
[
0.9281823039054871,
0.0008668478112667799,
0.07095090299844742
] |
[
0.9662076234817505,
0.03251732140779495,
0.0010886869858950377,
0.00018636361346580088
] |
en
| 0.999999 |
A general sense of perceived self-efficacy was assessed with the ten-item General Self-Efficacy Scale (GSE). 23 , 24 Items were rated on a four-point Likert scale ranging from 1 = ‘not at all true’ to 4 = ‘exactly true’, with a sum score ranging from 10 to 40. Higher values indicate higher levels of self-efficacy.
|
39494847_p12
|
39494847
|
Social and emotional aspects
| 2.64497 |
biomedical
|
Study
|
[
0.5843570828437805,
0.0008793374872766435,
0.41476356983184814
] |
[
0.965877890586853,
0.03314262628555298,
0.0008024160633794963,
0.00017704456695355475
] |
en
| 0.999997 |
Perceived stress was assessed using the Perceived Stress Scale (PSS-4), 25 , 26 which has four items on a five-point Likert scale ranging from 0 = ‘never’ to 4 = ‘very often’. The total sum score ranges from 0 to 16, with higher scores indicating more perceived stress.
|
39494847_p13
|
39494847
|
Social and emotional aspects
| 2.888902 |
biomedical
|
Study
|
[
0.9836720824241638,
0.0005521153216250241,
0.015775876119732857
] |
[
0.9862430691719055,
0.013070021755993366,
0.0005704823415726423,
0.00011638061550911516
] |
en
| 0.999998 |
Social support was assessed with the ESSI. 27 The five items were rated on a five-point Likert scale from 1 = ‘none of the time’ to 5 = ‘all of the time,’ with a total sum score ranging from 5 to 25. Higher scores indicate higher levels of social support.
|
39494847_p14
|
39494847
|
Social and emotional aspects
| 2.294055 |
biomedical
|
Study
|
[
0.91336590051651,
0.0014845671830698848,
0.08514951169490814
] |
[
0.964070200920105,
0.03469367325305939,
0.0009929590160027146,
0.00024312929599545896
] |
en
| 0.999999 |
The Brief Resilience Scale (BRS) 28 , 29 was used to measure the ability to bounce back or adapt well in the face of adversity (e.g. ‘I tend to bounce back quickly after hard times.’), with a five-point Likert scale ranging from 1 = ‘strongly disagree’ to 5 = ‘strongly agree’. The total sum score ranged from 1 to 5. Higher values indicate higher levels of resilience.
|
39494847_p15
|
39494847
|
Social and emotional aspects
| 2.405548 |
other
|
Study
|
[
0.45210739970207214,
0.001080421730875969,
0.546812117099762
] |
[
0.62203449010849,
0.3744213283061981,
0.0028181488160043955,
0.0007260082056745887
] |
en
| 0.999997 |
Participants in both samples were asked to rate their attitudes towards vaccination in general (‘What is your attitude towards vaccinations in general?’) on a five-point Likert scale from 1 = ‘rejecting’ to 5 = ‘supporting’.
|
39494847_p16
|
39494847
|
Attitude towards vaccination
| 1.771287 |
biomedical
|
Study
|
[
0.8021254539489746,
0.003384610405191779,
0.19448991119861603
] |
[
0.8814317584037781,
0.11619928479194641,
0.0017183420713990927,
0.0006505551282316446
] |
en
| 0.999995 |
Descriptive statistics on sociodemographic characteristics, depressive symptoms, social and emotional aspects, and attitudes towards vaccination in both samples (Brazilian and German university students) are reported. Chi-squared tests were performed to estimate the differences between the samples with respect to the following categorical variables: gender, relationship status, being a parent, residential status, study programme, migration background, clinically relevant depressive symptoms and suicidal thoughts. Where needed, significant effects found using χ 2 -tests were further decomposed using a z -test to compare column proportions. For group comparisons of the continuous dependent variable age, a t -test was performed.
|
39494847_p17
|
39494847
|
Statistical analysis
| 3.969906 |
biomedical
|
Study
|
[
0.9986634254455566,
0.00035884036333300173,
0.00097774644382298
] |
[
0.9995331764221191,
0.00020876069902442396,
0.0002196429413743317,
0.000038455764297395945
] |
en
| 0.999996 |
Separate one-way analyses of covariance (ANCOVAs) were performed to estimate the differences between Brazilian and German university students with respect to the continuous dependent variables (depressive symptoms, alcohol use, loneliness, social support, self-efficacy and attitude toward vaccinations). The mean scores were controlled for age and gender as covariates in all analyses. A bootstrapping procedure (1000 resamplings; 95% bias-corrected and accelerated confidence intervals (CI BCa)) was used to correct for group size differences and deviations from a normal distribution and to present 95% confidence intervals for the means. Bonferroni correction was applied to adjust for multiple testing when applicable. To estimate effect sizes for chi-squared tests, the ϕ coefficient was used; Cramér's V (ϕ c ) was used when the contingency table was larger than 2 × 2, with ϕ, ϕ c = 0.10 indicating a small effect, ϕ, ϕ c = 0.30 an average effect and ϕ, ϕ c = 0.50 a large effect. 30 The effect size was interpreted as small when d < 0.20, medium when d < 0.50 and large when d > 0.50 for the t -test; whereas η ²partial = 0.001 was interpreted as small, η ²partial = 0.06 as medium and η ²partial = 0.14 as large for the ANCOVAs. 30
|
39494847_p18
|
39494847
|
Statistical analysis
| 4.114481 |
biomedical
|
Study
|
[
0.9984754920005798,
0.00030254837474785745,
0.001221976475790143
] |
[
0.9993977546691895,
0.00021243216178845614,
0.0003550488909240812,
0.000034752094506984577
] |
en
| 0.999997 |
The analyses of sociodemographic data were used to present sample characteristics (i.e. analysis of potential sociodemographic differences between the samples of German and Brazilian students), whereas the analyses of psychosocial, emotional and behavioural outcomes were conducted to answer the research questions. All statistical analyses were performed using IBM SPSS Statistics version 27.0. A two-tailed α = 0.05 was applied to statistical testing.
|
39494847_p19
|
39494847
|
Statistical analysis
| 2.19099 |
biomedical
|
Study
|
[
0.963839590549469,
0.0010567402932792902,
0.03510371223092079
] |
[
0.9938355088233948,
0.005549571942538023,
0.00044606500887311995,
0.0001688706543063745
] |
en
| 0.999997 |
The Federal University of Parana's ethics committee and the Medical Faculty of Leipzig University's ethics committee granted approval for the current study. All participants provided informed consent before participation via an online opt-in function. This manuscript is an honest, accurate and transparent account of the study being reported, and no important aspects of the study have been omitted.
|
39494847_p20
|
39494847
|
Ethics statement
| 1.070019 |
other
|
Other
|
[
0.36544665694236755,
0.002348255831748247,
0.632205069065094
] |
[
0.009798646904528141,
0.9893442392349243,
0.0004691104986704886,
0.00038800601032562554
] |
en
| 0.999998 |
The sociodemographic characteristics of both samples are presented in Table 1 . There were statistically significant differences between Brazilian and German students for all sociodemographic variables. Specifically, German students were significantly more likely than Brazilian students to be female, single, not a parent, living alone and enrolled in a master's programme and to have a migration background. Brazilian students were older [mean deviation ( MD ) = 4.132, P < 0.001, 95% CI BCa (3.773, 4.516)], with ages ranging from 18 to 71 years ( M = 27.84, 95% CI BCa 27.49–28.22), compared with German students, whose ages ranged from 18 to 51 years old ( M = 23.71, 95% CI BCa (23.56, 23.85). Age was the only variable showing a large effect size ( d = 0.67). Table 1 Sample characteristics and group differences in sociodemographic information and health status Variable, n (%) a Brazilian students German students Test P -value Effect size Gender, n (%) χ 2 = 16.313 <0.001 ϕ c = 0.045 Female 1579 (64.8)* 3775 (69.0)* Male 821 (33.7)* 1599 (29.2)* Diverse 37 (1.5) 100 (1.8) Age, M (s.d.) 27.84 (8.41) 23.71 (4.81) t = 22.659 <0.001 d = 0.672 Relationship status, n (%) χ 2 = 29.527 <0.001 ϕ = 0.061 In a relationship 1349 (55.4) 2668 (48.7) Single 1088 (44.6) 2806 (51.3) Being a parent, n (%) 335 (13.7) 316 (5.8) χ 2 = 141.968 <0.001 ϕ = 0.134 Residential status, n (%) χ 2 = 137.505 <0.001 ϕ = 0.132 Alone 319 (13.1) 1355 (24.8) Not alone 2118 (86.9) 4119 (75.2) Study programme, n (%) χ 2 = 936.057 <0.001 ϕ c = 0.344 Bachelor 1404 (57.6)* 1928 (35.2)* Master 506 (20.8)* 3108 (56.8)* Other 527 (21.6)* 438 (8.0)* Migration background, n (%) χ 2 = 216.666 <0.001 ϕ c = 0.165 Self 28 (1.1)* 338 (6.2)* Parents 27 (1.1)* 362 (6.6)* None 2437 (97.7)* 4774 (87.2)* Diagnosed mental disorder 887 (36.5) 1239 (22.6) χ 2 = 164.008 <0.001 ϕ = 0.144 Chronic somatic diseases 383 (15.8) 933 (17.0) χ 2 = 1.811 0.178 ϕ = 0.178 Fully vaccinated against COVID-19 2372 (97.3) 5102 (93.2) χ 2 = 55.075 <0.001 ϕ = 0.083 Bold font indicates statistical significance. a. Percentages were calculated based on valid cases.
|
39494847_p21
|
39494847
|
Results
| 4.080198 |
biomedical
|
Study
|
[
0.9657257795333862,
0.0006642277585342526,
0.03360996022820473
] |
[
0.999481737613678,
0.0002948985493276268,
0.0001924687239807099,
0.000031012554245535284
] |
en
| 0.999997 |
Subsets and Splits
SQL Console for rntc/test-pp-aa
The query retrieves a sample of documents that are clinical cases with an educational score above 3, providing limited analytical value.
Clinical Cases Sample
Returns a sample of 100 clinical case documents, providing a basic overview of the document type's content.