TVnet: automated global analysis of tricuspid valve plane motion in CMR long-axis cines with residual neural networks for assessment of right ventricular function

Funding Acknowledgements

Type of funding sources: Foundation. Main funding source(s): Clarendon Fund, John Fell Oxford University Press Research Fund, Oxford BHF Centre of Research Excellence (RE/18/3/34214), Alison Brading Memorial Graduate Scholarship in Medical Science, National Institute for Health Research Oxford Biomedical Research Centre, National Institutes of Health (R01HL144706).

Background

Right ventricular (RV) function evaluation is an integral part of comprehensive cardiac assessment, including for pulmonary hypertension, congenital heart disease and arrhythmogenic RV cardiomyopathy (ARVC) [1]. It is commonly assessed by measuring tricuspid annular plane systolic excursion (TAPSE) and peak systolic velocity (RV s’) on echocardiography [2]. However, it is highly sensitive to imaging window and small changes in the beam angle, limiting reliability [3]. Cardiovascular magnetic resonance (CMR) is the imaging gold-standard for assessing RV structure and function, and is highly reproducible. CMR can assess tricuspid valve (TV) motion using four-chamber (4Ch) and RV two-chamber (2Ch) cines, with high diagnostic performance when compared against single-plane analysis [4]. However, manual placement of the TV insertion points is highly time-consuming for routine clinical workflows. TVnet, a deep-learning framework for automatically tracking the TV in 4Ch cines [5] has been recently validated, but without the orthogonal plane (RV 2Ch) which is helpful to more reliably characterise TV motion.

Purpose

We further extend TVnet to automatically track RV 2Ch cines and derive analysis of global TV motion parameters (global TAPSE and RV s’) on par with expert level performance.

Methods

74 patients undergoing CMR (1.5T Siemens MR scanner) with 4Ch and RV 2Ch views were retrospectively included in this ethically-approved study. The patients had the following cardiovascular conditions: myocardial infarction (n=43), ARVC (n=28) and Takotsubo cardiomyopathy (n=3). The dual-stage deep-learning pipeline with a residual neural network backbone [5, 6] (Figure 1A) was trained using 69 patient datasets and 15 patients were randomly chosen for testing. The TVnet trained on 4Ch cines [5] was used to automatically annotate the 4Ch cines of the testing set for global analysis comparison. For manual reference, the software Segment [7] was used to manually annotate the TV insertion points in all imaging data (1,865 RV 2Ch images, 375 4Ch images). The global TAPSE and RV s’ were derived as the mean perpendicular motion from the end-diastolic plane from both chamber views (Figure 1B).

Results

TVnet achieved a fast processing accuracy (<1 second per cine) and successfully tracked the TV insertion points from the RV 2Ch cine with a plane tracking error of 0.08±0.83 mm. The integrated pipeline yielded an excellent clinical-metric agreement with the manual reference (Figure 2) for both global TAPSE (error=0.42±0.68 mm, ICC=0.95) and RV s’ (error=-0.10±0.71 cm/s, ICC=0.85).

Conclusion

TVnet demonstrated excellent performance in both tracking the TV insertion points in RV 2Ch cines and deriving global TAPSE and RV s’ compared to manual reference. TVnet can eventually provide a complete automatic inline analysis of TV plane motion for a fast, reliable and reproducible assessment of RV function in routine clinical workflows. (A) TVnet pipeline (B) Metric derivationCorrelation and Bland-Altman plots

Ablation scar in a single pulmonary vein causes proarrhythmic mechanical destabilization in healthy sheep atria

Funding Acknowledgements

Type of funding sources: Public hospital(s). Main funding source(s): Catharina Hospital, Eindhoven Medtronic (unrestricted research grant)

Background

Ablative pulmonary vein isolation (PVI) prevents AF in 60% of AF patients. The absence of an antiarrhythmic effect of PVI is poorly understood. Atrial and PV stretch is proarrhythmic but the mechanical effect of PV ablation scar on AF arrhythmogenesis is unknown. We hypothesize that single ablation scars are potentially proarrhythmic because they create heterogeneous stretch.

Purpose

To evaluate the mechanical effect of a purposely incomplete PVI ablation scar on left atrial (LA) electrophysiology.

Methods

Functional cardiac MRIs in vivo in sheep (n = 11) before and 3-months after incomplete PVI by radiofrequency in the right PV (RPV) were analyzed with a feature-tracking algorithm to obtain local strain in the LA. The ablated hearts were explanted and perfused with 1:5 blood:Krebs solution in a dual-chamber working-heart set-up. Diagnostic multi-electrode endocardial catheters were positioned in the RPV and left PV (LPV). Premature stimulation was performed in each PV in low (∼12mmHg) and high (∼25mmHg) LA pressure. Twelve control hearts without ablation scar underwent similar ex vivo investigation.

Results

The maximum longitudinal strain of the myocardial wall between the RPV and LPV increased

 from 20.2 ± 6.2% to 33.5 ± 16.0% (before vs. after ablation, respectively; p = 0.032), whereas the maximum radial strain of the LA septum close to the RPV decreased from 45.6 ± 9.7% to 35.8 ± 7.3% (before vs. after ablation, respectively; p = 0.035). Sustained AF (>30s) was more often induced during stimulation in hearts with ablation scar than in control (25.0% and 11.5% of induction attempts (n = 76 and n = 87) in ablated and control hearts, respectively; p = 0.025). In ablated hearts, an increase in LA pressure augmented AF inducibility (12.8% vs. 37.8% of induction attempts (n = 39 vs. n = 37), low vs. high LA pressure, respectively; p = 0.023), whereas this was not the case in control hearts (4.4% vs. 19.0% of induction attempts (n = 45 vs. n = 42), low vs. high LA pressure; p = 0.289). The number of spontaneous premature atrial complexes (PACs) not leading to AF were similar in ablated and control hearts (0 ± 0 vs. 0 ± 2 total PACs within 20ms of refractory period during premature stimulation protocol, respectively; p = 0.411). The diastolic stimulation threshold of RPV was higher in the ablated than in control hearts (90 ± 63 vs. 79 ± 31mA, respectively; p = 0.049). The refractory period was similar in the ablated and control hearts (237 ± 62 vs. 235 ± 55ms, respectively; p = 0.873).

Conclusion

Local ablation scar caused regionally disparate bio-mechanical changes in proximity to ablative energy delivery and increased inducibility of sustained AF especially during increased LA stretch. This was associated with decreased tissue excitability without changes in refractoriness. A single incomplete PVI ablation scar therefore is proarrhythmic. Development of ablation lesion sets that homogenize atrial mechanics and electrophysiology may improve AF ablation success.

T1 mapping-derived signature of myocardial involvement in idiopathic inflammatory myopathy compared to acute viral myocarditis: a texture-based analysis

Funding Acknowledgements

Type of funding sources: None.

Background. Recent studies revealed the ability of MRI T1 mapping to characterize myocardial involvement in both idiopathic inflammatory myopathy (IIM) and acute viral myocarditis (AVM), as compared to healthy controls. However, neither myocardial T1 nor T2 maps were able to discriminate between IIM and AVM patients, when considering conventional myocardial mean values and derived indices such as lambda and extracellular volume.

Purpose. To investigate the ability of T1 mapping-derived texture analysis to differentiate IIM from AVM.

Methods. Forty patients, 20 with IIM (51 ± 17 years, 9 men) and 20 with AVM (34 ± 13 years, 16 men) underwent 1.5T MRI T1 mapping using a modified Look-Locker inversion-recovery sequence before and 15 minutes after injection of a gadolinium contrast agent. After manual delineation of endocardial and epicardial borders and co-registration of all inversion time images, native and post-contrast T1 maps were estimated. Myocardial texture analysis was performed on native T1 maps. Textural features such as: autocorrelation, contrast, dissimilarity, energy and sum entropy were used to build a least squares-based linear regression model. Finally, receiver operating characteristic (ROC) analysis was used to investigate the ability of such texture features score to classify IIM vs. AVM patients, compared to the performance of mean myocardial T1. A Wilcoxon rank-sum test was also used to test difference significance between groups.

Results. Both native and post-contrast mean myocardial T1 values were comparable between IIM (native: 1022 ± 43 ms; post-contrast: 319 ± 44 ms) and AVM (1056 ± 59 ms, p = 0.07; 318 ± 35 ms, p = 0.90, respectively) groups. Results of ROC analyses are provided in the Table, indicating that a better discrimination between IIM and AVM patients was obtained when using texture features, with higher AUC and accuracy than mean T1 values (Figure).

Conclusion. Texture analysis derived from MRI T1 maps without contrast agent injection was able to discriminate between IIM and AVM with higher accuracy, sensitivity and specificity than conventional T1 indices. Such analysis could provide a useful myocardial signature to help diagnose and manage cardiac alterations associated with IIM in patients presenting with myocarditis and primarily suspected of AVM.

Table Area under curve (AUC) Accuracy Sensitivity Specificity Native T1 0.67 0.70 0.65 0.75 Post-contrast T1 0.49 0.60 0.25 0.95 Texture features score 0.85 0.82 0.90 0.75 ROC analyses for classification between IIM and AVM patients Abstract Figure

Capturing complex right ventricular wall motion abnormalities in arrhythmogenic right ventricular cardiomyopathy by combining longitudinal and radial myocardial dynamics in feature-tracking MRI

Funding Acknowledgements

Type of funding sources: Public Institution(s). Main funding source(s): INSERM Liliane Bettencourt doctoral grant

Background

Arrhythmogenic Right Ventricular Cardiomyopathy (ARVC) is ass­­ociated with complex spatial and temporal right ventricular (RV) wall motion abnormalities. While cardiac magnetic resonance (CMR) is the gold-standard imaging technique, its diagnosic performance remains suboptimal and additional CMR biomarkers reflecting ARVC pathophysiology are needed.

Purpose

To evaluate the performance of a CMR feature-tracking (FT)-derived parameter combining both longitudinal and radial RV deformation and motion for the characterization of RV wall motion abnormalities in ARVC.

Methods

Thirty-nine patients with definite or borderline ARVC (median age 45 years, interquartile range 31–51, 56% males) were compared to 20 healthy controls with comparable age, sex and weight distributions. All subjects had 1.5T CMR including short axis and 4-chamber views steady-state free precession acquisitions. A custom FT software adapted to RV wall segmentation and tracking was used to assess RV wall deformation and motion in the 3 space directions resulting in: 1) global longitudinal strain (GLS) estimated on the 4 chamber view from the RV free wall, 2) basal circumferential strain (BCS) and radial motion fraction (BRMF) estimated as an average of short-axis slices comprised in the RV third basal portion. To capture the complex RV motion in ARCV, a longitudinal to radial strain loop (LRSL) was displayed and its area was calculated.

Results

The ARVC group comprised 28 (72%) patients with definite and 11 (28%) with borderline diagnosis . As compared to controls, LVEF and RVEF were significantly lower in ARVC patients (61(interquartile range (IQR) 52-71) vs. 71%(IQR 55-88) , p = 0.03 and 47%(IQR 16-63) vs. 57%(IQR 49-63) , p = 0.02, respectively), LVEF remaining within normal range limits. While there was no significant difference in RV GLS between ARVC patients and controls (median -17.7%(IQR -24–15) vs. -17.5%(IQR -20.1–15.2), p = 0.67) , BCS and BRMF were significantly lower in ARVC patients vs. controls [-7.5%(IQR -12.3–8.4.) vs. -9.8%(IQR -13.8–8.6.), p = 0.004 and -12.2(IQR -14.4–8.7.) vs. -14.9%(IQR -16.6–13.2) p = 0.0007, respectively] . The LRSL area was significantly and markedly lower in ARVC patients vs. controls [70.6 (IQR 16.3-63.1) vs. 144.1 (IQR 110.4-251.3), p = 0.0002] . LRSL area outperformed RVEF, BCS and BRS in separating ARVC from controls (area under receiving operator characteristics curve 0.82 vs. 0.78, 0.73 and 0.78, respectively).

Conclusion

In ARVC, a FT-derived parameter combining longitudinal and radial RV wall deformation and motion provided better discrimination of ARVC patients from controls than conventional FT measurements. Its implementation in clinical practice may bolster CMR performance to characterize ARVC wall motion abnormalities.

Abstract Figure

Wave decomposition applied to LA phasic longitudinal strain evaluated from MRI feature tracking to estimate a true LA booster strain index

Funding Acknowledgements

Type of funding sources: None.

Background. Feature tracking (FT) is an emerging approach for the evaluation of both left atrium (LA) and left ventricular (LV) myocardial strain from the same cine MRI dataset. We hypothesized that the LA active contraction longitudinal strain, is a merge of an intrinsic LA booster contraction with the early diastolic LA emptying, especially when this latter is extended because of a poor LV relaxation (Figure 1, bottom). Such index can be estimated through LA phasic strain wave-decomposition as conventionally done for pressure curves to estimate forward and reflected components.

Purpose. To compare the newly proposed LA intrinsic or "true" booster index (Sla_fit) against the conventional index (Sla) in terms of associations with LV remodeling (LV mass/ LV volume), LV systolic longitudinal strain (LV_GLS), and transmitral LV filling indices in healthy controls and aortic valve stenosis (AVS) patients with preserved LV ejection fraction.

Methods. We studied 55 patients (34 AVS:71 ± 11years, 21 controls:66 ± 9years) who had an MRI exam with cine SSFP and phase contrast (PC) images. FT was applied to cine images to extract LV and LA phasic longitudinal strain and strain rates. Transmitral flow early (E, cm/s) and late (A, cm/s) filling peak velocities were calculated from PC data. To estimate intrinsic LA booster index, the LA longitudinal strain curve corresponding to the reservoir and conduit phases was fitted using two half cosine waves, to account for an eventual LA filling to LA early emptying asymmetry, while fitting the LA contraction with a full cosine wave (Figure 1). The peak of this latter wave was defined as the intrinsic LA booster strain index (Sla_fit), while the second peak of the measured LA strain was defined as the conventional LA booster strain (Sla).

Results. While conventional Sla was significantly higher than intrinsic LA booster Sla_fit in AVS patients (13.55 ± 4.26 vs. 8.09 ± 6.07, p = 0.0002), it was nearly equivalent in controls (14.34 ± 4.30 vs.13.43 ± 4.23, p =.49). But the newly proposed LA booster strain index was significantly related to LV_GLS (r=-48,p=.0004); to LV remodeling (r=-.44,p = 0.0012) as well as to transmitral flow A wave ( r=-.49, p=.0005) none of these associations were significant when considering conventional LA booster strain. Interestingly our intrinsic LA booster index Sla_fit was significantly associated with LV longitudinal strain in both controls (r=-.55,p = 0.009) and asymptomatic AVS (N = 10) (r=-.77,p = 0.0081) but not in symptomatic AVS (N = 24) (p>.70). This may reveal a maintained LA-LV coupling in the asymptomatic phase and an uncoupling in the symptomatic phase, caused by elevated LV filling pressures.

Conclusions. A promising index for the quantitative evaluation of intrinsic LA booster function was proposed and its consistency was demonstrated through its significant associations with LV remodeling, LV longitudinal strain and transmitral late filling peak.

Abstract Figure.

161 EFFECT OF OVIDUCTAL FLUID ON BOVINE OOCYTE ZONA PELLUCIDA HARDENING AND SPERM BINDING, AND ON EARLY EMBRYO DEVELOPMENT

This study was designed to evaluate the effects of oviductal fluid at different periovulatory times on oocyte maturation, modification of the zona pellucida (ZP), fertilization and embryo development. Bovine oviducts were collected at a slaughterhouse and classified as preovulatory (pre-ov: 1 pre-ov follicle and a regressing corpus luteum) or post-ovulatory (post-ov: a corpus haemorrhagicum or recent corpus luteum; n = 10 cows/stage). Both oviducts were flushed with 1 mL of sterile TCM-199, and oviductal flushes (OF) were aliquoted and stored at –80°C. Abattoir-derived bovine ovaries were aspirated and cumulus‐oocyte complexes (COC) with at least 3 cumulus layers and homogeneous oocyte cytoplasm were in vitro matured for 22 h in standard maturation medium (control group, n = 319) or in standard medium with 2× concentrated additives supplemented (50% v/v) with pre-ov OF (n = 255) or post-ov OF (n = 248). After in vitro maturation (IVM), subgroups of COC were denuded, and the time of digestion of the ZP by pronase 0.1% (v/v in TCM-199) was determined to evaluate ZP hardening. After IVM, COC were fertilised in vitro for 18–20 h at a final concentration of 1.106 million spermatozoa (spz)/mL. After in vitro fertilization (IVF), COC were denuded, washed twice and cultured for 8 days more under standard conditions. After IVM, IVF, and embryo culture, oocytes/embryos were fixed with ethanol, stained with Hoescht, and examined under fluorescence microscopy for determination of (1) maturation and developmental stages, (2) numbers of fertilised and polyspermic oocytes, and (3) spz bound to the ZP. Percentages were compared between groups by chi-square. Times of ZP digestion were compared by Kruskal‐Wallis test. Numbers of spz bound to the ZP were compared by ANOVA on normalised data followed by Newman-Keuls tests. Data are presented as mean ± SEM. A P < 0.05 was considered significant. Addition of OF during IVM had no effect on maturation rates compared with the control. However, the digestion time of the ZP by pronase was reduced after IVM with pre-ov OF (313 ± 21 s; n = 26) compared with post-ov OF (459 ± 23 s; n = 23) but not with the control (416 ± 30 s; n = 25). After IVF, the number of spermatozoa bound to the ZP was increased after IVM with pre-ov OF (57 ± 5 spz/oocyte; n = 67) and decreased after IVM with post-ov OF (34 ± 3 spz/oocyte; n = 76) compared with the control (42 ± 5 spz/oocyte; n = 60). Addition of OF during IVM had no effect on rates of IVF and polyspermia. However, the rate of development to the blastocyst stage was less after IVM with post-ov OF (10%, n = 97 cleaved oocytes) compared with control (24%, n = 130) and pre-ov OF (29%, n = 101). In conclusion, the OF collected before ovulation decreased the resistance of the ZP to protease digestion and increased its ability to bind spz, whereas it was the opposite for the post-ov OF. Furthermore, the post-ov OF decreased the developmental competence of fertilised oocytes.

74 ANALYSIS OF STEROID HORMONES IN BOVINE OVIDUCTAL FLUID BY GAS CHROMATOGRAPHY COUPLED WITH TANDEM MASS SPECTROMETRY

Steroid hormones play key roles in the regulation of physiological changes in the mammalian genital tract, including the oviduct. The aim of this study was to determine the variations in steroid hormone concentrations in bovine oviducal fluid (OF) during the oestrous cycle. Bovine oviducts were collected at a local slaughterhouse and classified into 4 stages according to the ovarian and corpus luteum (CL) morphologies (n = 18–25 cows/stage): post-ovulatory (postov), mid luteal (mid-lut), late luteal (late-lut), and preovulatory (preov). Follicular fluid was also collected from the preov follicles. Animals with follicular oestradiol-17β (E2) <40 ng mL–1 or progesterone (P4) >160 ng mL–1 (cystic follicles) or E2 : P4 ratio <1 (atretic follicles) were excluded. Oviducal fluids were collected from contra- and ipsilateral (to the CL) ampullas by squeezing. They were then centrifuged and stored at –80°C. Steroids were extracted from pools of 150 to 200 µL of OF (4–10 cows/pool; 3–4 pools per “stage × side” group), purified, fractioned by high-performance liquid chromatography, derivatized, and analysed by gas chromatography coupled with tandem mass spectrometry (GC-MS/MS). Comparisons between groups were made with a two-way ANOVA followed by Bonferroni post-tests. Data are expressed as means ± SEM. A P-value < 0.05 was considered as significant. In ipsilateral OF, the concentrations of P4 increased from postov (56.9 ± 13.4 ng mL–1) to mid-lut (120.3 ± 34.3 ng mL–1; P < 0.01) then decreased from late-lut (76.7 ± 1.8 ng mL–1) to preov (6.3 ± 1.7 ng mL–1; P < 0.001) stages, and were 4 to 16 times more concentrated than in contralateral OF. Several P4 reduced metabolites (5α-dihydroP4, 20α-dihydroP4, 17α-hydroxyP4, 3α5α20α-hexahydroP4, 3β5α20α-hexahydroP4) followed the same pattern of variation. Concentrations of the P4 precursor pregnenolone were highest at postov (4.2 ± 0.3 ng mL–1 in ipsilateral OF) then decreased from mid-lut to preov (3.4 ± 0.6 to 1.3 ± 0.1 ng mL–1; P < 0.001) and were higher than in contralateral OF at postov (P < 0.001) and mid-lut (P < 0.01) stages. Concentrations of E2 were higher in preov (290.5 ± 63.2 pg mL–1 in ipsilateral OF) compared to all other stages (P < 0.05) with no difference between ipsi- and contralateral sides at any stage. Concentrations of the oestrogen precursors androstenedione and testosterone displayed the same pattern of variation as E2. In contrast, concentrations of dehydroepiandrosterone and estrone did not vary between stages or sides relative to the CL. In addition, high concentrations of cortisol and cortisone were found in ipsi- and contralateral OF at all stages of the oestrous cycle (on average 48.6 and 31.9 ng mL–1, respectively). In conclusion, numerous steroids, steroid precursors, and metabolites were found in the bovine OF. The local transfer of ovarian steroids creates a highly concentrated and fluctuating hormonal environment in the ipsilateral OF during the oestrous cycle. These results could be useful to improve media for IVF, development, and oviducal cell culture.

A custom robot for Transcranial Magnetic Stimulation: first assessment on healthy subjects

In this paper, a custom robotic system for Transcranial Magnetic Stimulation is assessed in clinical conditions on healthy subjects. A motor cortex mapping is performed using the robotic system with comparison to a manual approach using a neuronavigation system. Stimulation accuracy, repeatability are evaluated as well as the feeling of the system operator and the subject in terms of comfort, tiredness, stress level, ease-of-use. Very encouraging results are obtained on all these aspects, which strengthens the idea of developing robotic assistance for TMS.

White matter volume is decreased in the brain of patients with neuromyelitis optica

BACKGROUND

Neuromyelitis optica (NMO) is an inflammatory disease involving predominantly the spinal cord and optic nerves. Whether patients with NMO have a loss in white or grey matter (GM) volumes remains to be determined.

METHODS

Thirty patients with NMO, 30 healthy subjects matched for age and gender, 21 patients with multiple sclerosis (MS) and 20 patients with a clinically isolated syndrome (CIS) were studied. We applied a SIENAX post-treatment software. We compared white matter (WM) and GM volumes between groups and explored correlations of changes in NMO patients with age, gender, duration, disease severity, visual acuity and T2 hyperintensities. We also performed a voxel-based morphometry (VBM) analysis to identify the regions affected by loss of volume.

RESULTS

White matter volume was significantly reduced in patients with NMO (764.4 ± 58.3 cm(3) ) compared to healthy subjects (843.1 ± 49.3 cm(3) ) (P < 0.001), whereas no difference was observed for the GM. Patients with CIS also presented an elective atrophy of WM and MS an atrophy of both WM and GM. We did not find any predictive factors of brain atrophy. The decrease in WM volume in NMO was noted even in the absence of visible MRI hypersignals. The VBM analysis found a few regions of WM atrophy (corpus callosum and optic radiations, P < 0.005, uncorrected) and a few regions of GM atrophy (thalamus and prefrontal cortex, P < 0.001, uncorrected).

CONCLUSION

These results suggest a significant brain involvement in NMO, especially an involvement of WM which appears not to be limited to secondary degeneration after spinal cord and optic nerve damage.

Reassembly of Lumbricus terrestris hemoglobin: a study by matrix-assisted laser desorption/ionization mass spectrometry and 3D reconstruction from frozen-hydrated specimens

Dodecamers and four types of linker chains (L1-L4) were purified from dissociated hemoglobin of the earthworm Lumbricus terrestris. Various preparations comprising dodecamer of globin chains and linker chains were allowed to reassemble at neutral pH. They produced various oligomers that were purified by gel filtration, analyzed in matrix-assisted laser desorption/ionization mass spectrometry and submitted to 3D reconstruction from isolated particles observed in cryoelectron microscopy. Despite the impossibility to completely free the L2, L3, and L4 preparations from L1, the following conclusions were obtained. First, hemoglobin molecules indistinguishable from native hemoglobin at 25 A resolution were obtained in the absence of linker chains L2, L3, or L4. Second, the 3D reconstruction volumes of reassembled hemoglobins containing dodecamers and L1+L3 or dodecamers and L1+L4 demonstrate that reassembly of native-like structures can be obtained from at most two linker chains and dodecamers. Third, the 3D reconstruction volumes of native and reassembled hemoglobins containing dodecamers and (1) L1, L2, and L4, (2) L1, L3, and L4, (3) L1 and L4, and (4) L1 and L3 were highly similar. Since these structures comprise two types of substructures (one involved in the c3a, c3b, and c4 linking units of the hollow globular substructure and the other in the c5 connection and the toroid), it seems highly probable that the minimal number of linker chains required to reassemble native-like hemoglobin is at most two.

Octamerization enables soluble CD46 receptor to neutralize measles virus in vitro and in vivo

A chimeric fusion protein encompassing the CD46 ectodomain linked to the C-terminal part of the C4b binding protein (C4bp) alpha chain (sCD46-C4bpalpha) was produced in eukaryotic cells. This protein, secreted as a disulfide-linked homo-octamer, was recognized by a panel of anti-CD46 antibodies with varying avidities. Unlike monomeric sCD46, the octameric sCD46-C4bpalpha protein was devoid of complement regulatory activity. However, sCD46-C4bpalpha was able to bind to the measles virus hemagglutinin protein expressed on murine cells with a higher avidity than soluble monomeric sCD46. Moreover, the octameric sCD46-C4bpalpha protein was significantly more efficient than monomeric sCD46 in inhibiting virus binding to CD46, in blocking virus induced cell-cell fusion, and in neutralizing measles virus in vitro. In addition, the octameric sCD46-C4bpalpha protein, but not the monomeric sCD46, fully protected CD46 transgenic mice against a lethal intracranial measles virus challenge.

Structural comparison of cephalopod hemocyanins: phylogenetic significance

Hemocyanins, the respiratory molecules of cephalopod mollusks, are hollow cylinders with five internal arches. Three hemocyanins representative of three orders of cephalopods (Benthoctopus species, Octopoda; Vampyroteuthis infernalis, Vampyromorpha; Sepia officinalis, Sepioidea) were subjected to cryoelectron microscopy and three-dimensional (3D) reconstruction. The structure of Benthoctopus hemocyanin, solved at 26.4-A resolution, possesses arches comprising two identical functional units. The similarity between these functional units and the structure recently observed in X-ray crystallography for Octopus by Cuff et al. (J. Mol. Biol., 1998, 232, 522-529) allows the identification of their N- and C-terminal domains in the 3D reconstruction volume. Conversely, arches present in the 3D reconstruction volume of Sepia hemocyanin (21.8 A resolution) contain four functional units that are disposed differently. The strong resemblance between the reconstruction volumes of Vampyroteuthis (21.4-A resolution) and Benthoctopus hemocyanins suggests that Sepioidea diverged from a group containing Octopoda and Vampyromorpha.

Intramolecular localization of the functional units of Sepia officinalis hemocyanin by immunoelectron microscopy

The quaternary structure of Sepia officinalis hemocyanin (Hc) as studied in immunoelectron microscopy with rabbit IgGs and Fab fragments raised against functional units (FU) Soc, Sod, Soe, Sof, Sog, and Soh and fragment Soab. The architecture of immunocomplexes shows that (i) epitopes characteristic of FUs Soc and Sog and of fragment Soab are located in the two external tiers of FUs, (ii) FUs Soh and Soe or Sod are located in arches. These results were confirmed using immunocomplexes made up of Sepia Hc and IgGs or Fab fragments purified from antisera raised against FUs of Octopus vulgaris and Octopus dofleini. Frozen-hydrated immunocomplexes containing one Hc molecule and at least one FU-specific Fab fragment were observed in the electron microscope and submitted to image processing. When the Hc molecule is viewed along its 5-fold axis (i) anti-Soc Fab fragments project on a radius passing through the arch's pillar, (ii) anti-Sof Fabs project slightly out of the arches, and (iii) anti-Soh Fabs project between neighboring arches. When applied to a recent three-dimensional (3D) reconstruction volume, these results allow us to deduce the intramolecular location of five of the eight FUs. For the last three FUs limited uncertainties remain: (i) Soc can be located in two positions in the external tier of FUs; (ii) Soa and Sob can both occupy three positions in the external tiers; and (iii) because of an immunological cross-reactivity Sod may be located in the wall and Soe in the arch, or vice versa. An analysis of the quaternary structure considering the possible locations of the 80 FUs and postulating a single type of subunit shows that 80 possibilities of paths still exist for the polypeptide chain. To solve definitely these 80 possibilities only five questions remain to be answered.

Lumbricus terrestris hemoglobin: a comparison of small-angle x-ray scattering and cryoelectron microscopy data

The quaternary structure of Lumbricus terrestris hemoglobin was investigated by small-angle x-ray scattering (SAXS). Based on the SAXS data from several independent experiments, a three-dimensional (3D) consensus model was established to simulate the solution structure of this complex protein at low resolution (about 3 nm) and to yield the particle dimensions. The model is built up from a large number of small spheres of different weights, a result of the two-step procedure used to calculate the SAXS model. It accounts for the arrangement of 12 subunits in a hexagonal bilayer structure and for an additional central unit of clylinder-like shape. This model provides an excellent fit of the experimental scattering curve of the protein up to h = 1 nm-1 and a nearly perfect fit of the experimental distance distribution function p(r) in the whole range. Scattering curves and p(r) functions were also calculated for low-resolution models based on 3D reconstructions obtained by cryoelectron microscopy (EM). The calculated functions of these models also provide a very good fit of the experimental scattering curve (even at h > 1 nm-1) and p(r) function, if hydration is taken into account and the original model coordinates are slightly rescaled. The comparison of models reveals that both the SAXS-based and the EM-based model lead to a similar simulation of the protein structure and to similar particle dimensions. The essential differences between the models concern the hexagonal bilayer arrangement (eclipsed in the SAXS model, one layer slightly rotated in the EM model), and the mass distribution, mainly on the surface and in the central part of the protein complex.

Three-dimensional reconstruction of Limulus polyphemus hemocyanin from cryoelectron microscopy

Hemocyanin (Hc) the respiratory pigment of the horseshoe crab Limulus polyphemus (Lp) is composed of 48 approximately 75 kDa copper-containing subunits arranged in eight hexameric groups. In this study, we used the random conical tilt series method to do a three-dimensional (3D) reconstruction of Lp Hc observed in vitreous ice. This approach allowed the unambiguous determination of the handedness of the molecule. Lp Hc contains two superimposed 4 x 6mer structures possessing the same structural features as the other 4 x 6meric Hcs, namely flip and flop views and a rocking effect. Moreover, 3D fitting of the X-ray structure of subunit LpII with the reconstruction volume shows that the intra4 x 6meric contacts described in arthropod Hcs also occur within Limulus Hc. The two half-molecules composing the 8 x 6mer have their flop faces in contact (flop/flop association), the main links being formed by subunits LpIV. Model building shows that the flop/flop association is the only possible arrangement which allows the assembly of the whole particle. The two alternate constructions (flip/flop and flip/flip) are forbidden because of steric hindrance.

Similar architectures of native and transformed human alpha2-macroglobulin suggest the transformation mechanism

The refined three-dimensional structure of native human alpha2-macroglobulin (alpha2M) has been determined by cryoelectron microscopy and three-dimensional reconstruction. New features corresponding to "sigmoid arches," "basal bodies," and "apical connections" were observed in the molecule. Since similar elements are found in the architecture of transformed alpha2M, the 2 volumes were aligned in three dimensions. In their common orientations, they show many similarities except near the openings of the central chamber. In the native conformation, these apertures are fully opened, allowing the proteases to access the central chamber of the molecule, while in the transformed structure, they are partially closed. These structures suggest that alpha2M conformational change involves a strong lateral compression and a vertical stretching of the native particle seen in its four-petaled flower view to produce the H view of the transformed form. A model of structural transformation, in which all the parts of the alpha2M molecule seem involved in the entrapment of the proteinases is proposed.

Complete amino acid sequence of the Aa6 subunit of the scorpion Androctonus australis hemocyanin determined by Edman degradation and mass spectrometry

The primary structure of the hemocyanin Aa6 subunit from the scorpion Androctonus australis was resolved by using protein sequencing and mass spectrometry for analysis of the polypeptide chain and of fragments obtained by CNBr, trypsin, and chymotrypsin cleavage. Due to the high sensitivity of the methodologies used, only a small amount of material, less than 1 mg, was consumed. The complete sequence is composed of 626 amino acid residues and the protein is not glycosylated but probably phosphorylated at Ser374. Its molecular mass measured by mass spectrometry (71,890 +/- 7 Da) is about 30 Da higher than the mass calculated from the sequence data (71,860.1 Da). The origin of this difference is not clear but could result from minor molecular heterogeneities. Within the chelicerates, the Aa6 subunit of the arachnid A. australis shares 405 identical residues with chain e of another arachnid, Eurypelma californicum, and 399 with chain alpha of the merostom Tachypleus tridentatus. The degrees of identity between these three subunits, which are known to occupy the same location in the native hemocyanin oligomers, are significantly higher than those existing between the subunits a, d, and e of E. californicum. This favors the hypothesis that gene duplications, leading to separate chains in one species, have occurred before the divergence between arachnids and merostoms.

Three-dimensional reconstruction of Androctonus australis hemocyanin labeled with a monoclonal Fab fragment

An immunocomplex formed by the 4 x 6 meric hemocyanin of the scorpion Androctonus australis with the monoclonal antibody L104 was studied by cryoelectron microscopy and subjected to three-dimensional (3D) reconstruction. The reconstructed particle reflects the structure of the immunocomplex in its hydrated state and is devoid of the flattening that was previously observed with a negatively stained preparation. A 3D fitting of the X-ray data of the Panulirus interruptus hemocyanin and of the Fab R19.9 to the reconstruction volume allowed the first quantitative measurement of the structural parameters of the antigen, and the localization of the epitope at the surface of subunit Aa6. The independent alignment of the Fabs and the hexamers provides a direct verification for the accuracy of the fit and allows the building of the most detailed model of a cheliceratan 4 x 6meric complex and its attached monoclonal antibodies.

Architecture of native human alpha 2-macroglobulin studied by cryoelectron microscopy and three-dimensional reconstruction

The architecture of the native human alpha 2-macroglobulin was studied by cryoelectron microscopy and image processing techniques. The lip, padlock, doughnut, and four-petaled flower views of this homotetrameric proteinase inhibitor were observed in the frozen-hydrated specimen, and a new view, termed eye view, was also characterized. The present three-dimensional reconstruction demonstrates that all these electron microscope views derive from a single three-dimensional structure. The molecule is composed of two horizontal bodies and of two oblique arches, which border a large central cavity. The polymorphism and the flexibility of the native alpha 2-macroglobulin are discussed.

Quaternary structure of Octopus vulgaris hemocyanin. Three-dimensional reconstruction from frozen-hydrated specimens and intramolecular location of functional units Ove and Ovb

A frozen-hydrated sample of Octopus vulgaris hemocyanin was imaged at 0 degree and 40 degrees tilt angle under low dose conditions by transmission electron microscopy. A three-dimensional reconstruction by the method of random conical tilt series produced a three-dimensional volume to which a D5 symmetry was applied. Examination of serial sections in the volume and surface representation at various thresholds allowed the five arches containing functional unit Ovg to be localized at the interdimeric subunit groove. In another set of experiments specific polyclonal antibodies were used to label functional units Ovb and Ove in the cylinder wall. The observation of the negatively stained immunocomplexes showed that Ovb is located in the external tiers of functional units and Ove in the internal tier. These results suggest that the direction of the polypeptide chains in the cylinder wall may be only partially antiparallel. A model of the quaternary structure is proposed with the following features: (1) the external tiers of functional units comprise four units each (Ova-d) coming from a single polypeptide chain; (2) the internal tier comprises two functional units from each polypeptide chain (Ove-f); (3) the interdimeric subunit arches connect the two copies of a single functional unit (Ovg) located in each polypeptide chain.

Three-dimensional immunoelectron microscopy of scorpion hemocyanin labeled with a monoclonal Fab fragment

An immunocomplex of the 4 x 6-meric hemocyanin of the scorpion Androctonus australis with the monoclonal Fab fragment L104 was reconstructed from electron micrographs of a negatively stained specimen, using the double-carbon-layer technique. The resulting structure enables a clear visualization of the Fab fragments bound to the four copies of the Aa6 subunit and directly confirms a previous localization of the L104 epitope deduced from two-dimensional image processing. Despite a strong flattening effect produced by the negative-staining technique the orientations of the Fab fragments are well characterized. Moreover, the observation of a central hole within the elbow bends of the Fab fragments provides information about the disposition of the Fabs around their main axis.

Further approaches to the quaternary structure of octopus hemocyanin: a model based on immunoelectron microscopy and image processing

The direction of the polypeptide chains and the location of the functional units in Octopus vulgaris hemocyanin were studied by various methods. Monoclonal antibodies specific for the Ovc (clone Ov409) and Ovg (clone Ov315) functional units produced immunocomplex strings which were examined in the electron microscope. In both cases the immunocomplexes contained more than two hemocyanin molecules in their side view, demonstrating that in the whole hemocyanin neighboring polypeptide chains run in antiparallel directions. The interhemocyanin distances in the immunocomplexes also indicated that Ovg is located inside the cylinder, while Ovc is located in the external layers of functional units. In addition, the fact that the binding point of the Fab arm to the hemocyanin molecule was occasionally visible confirmed the external location of functional unit Ovc. Image processing of the whole hemocyanin cross-linked with dimethyl suberimidate showed that the end-on view is not a perfect cylinder but a regular pentahedron and that the five-arch collar is probably composed of five pairs of functional unit Ovg located inside the cylinder. The accessibility of cross-linked hemocyanin to functional unit-specific polyclonal antibodies, studied in immunoelectrophoresis, showed that Ovb and Ove are highly accessible, while Ovd, Ovf, and Ovg are not. The low accessibility of Ovd may be at least partially explained by its high sugar content which could hamper the accessibility of the antibody to the antigen.

Three-dimensional architecture of human alpha 2-macroglobulin transformed with methylamine

A frozen-hydrated sample embedded in vitreous ice of human alpha 2-macroglobulin transformed by methylamine was imaged by cryoelectron microscopy and reconstructed in three dimensions. In the reconstruction, the cage-like architecture of this protease inhibitor is fully revealed with a clear visualization of two lozenge-shaped lateral walls connected by thin bridges. The shape and dimensions of the internal cavity normally containing the trapped protease(s) is described. The possible locations of the thiol ester sites and inter-subunit connections are also discussed.

Electron microscopy of alpha 2-macroglobulin with a thiol ester bound ligand

In order to covalently bind the hydrolyzed thiol ester groups of the human alpha 2-macroglobulin (alpha 2M) transformed by methylamine, the phospholipase A2 (PLA2), a small enzyme (M(r) = 13,000) from Naja nigricollis snake venom was activated by succinimidyl 4-(maleimidomethyl)cyclohexane-1-carboxylate (SMCC). Average images determined from electron micrographs of the methylamine-transformed alpha 2M, with and without activated PLA2, were determined by image processing and compared. A localization of the PLA2 was achieved by subtracting the average image of alpha 2M transformed by methylamine from that containing PLA2. The results are consistent with previous work showing the central localization of chymotrypsin trapped in alpha 2M. They also suggest that the four thiol esters are located near the center of the alpha 2M molecule.

Three-dimensional reconstruction of native Androctonus australis hemocyanin

A sample of native 4 x 6-meric hemocyanin of Androctonus australis was negatively stained with the double-layer technique, and was observed by transmission electron microscopy under low-dose conditions with a 50 degree and 0 degree tilt. The three-dimensional reconstruction method from "Single-exposure, random conical tilt series" was then applied. Independent three-dimensional reconstructions were obtained from the top, side and 45 degree views. Despite a pronounced flattening effect, presumably due to the specimen preparation technique, the positions of the 24 subunits composing the oligomer were unequivocally determined. This experiment definitely solves the problem of the architectural organization of the subunits in the cheliceratan 4 x 6-meric hemocyanins. Moreover, distinction between the flip and flop faces and an attenuated rocking effect were observed.

Crystals of the carboxyl-terminal functional unit from Octopus dofleini hemocyanin

The carboxyl-terminal oxygen-binding unit of the polypeptide from Octopus dofleini hemocyanin has been crystallized in a form suitable for three-dimensional X-ray analysis. This proteolytic fragment has a molecular weight of 47 kDa and reversibly binds O2 while exhibiting a slight Bohr effect. Two types of crystals have been grown. Type I crystals, currently under analysis, belong to the orthorhombic space group P2(1)2(1)2(1) and have unit cell dimensions of 92.6 A x 167.4 A x 59.2 A. A composition of two protein molecules per asymmetric unit and 50% solvent content is consistent with a self-rotation function that identifies a non-crystallographic 2-fold axis of symmetry relating these molecules. Diffraction extending beyond 1.9 A Bragg spacings can be detected with synchrotron X-radiation.

Topological mapping of 13 epitopes on a subunit of Androctonus australis hemocyanin

A topological localization of epitopes on the surface of the Aa6 subunit of Androctonus australis hemocyanin has been carried out. First, immunocomplex strings composed of native hemocyanin and monoclonal antibodies were examined in the electron microscope and submitted to an image processing by correspondence analysis. The average images were then compared to a three-dimensional model of the 24-mer suggesting that 11 of the 13 epitopes are located in three zones of the subunit surface. Second, the overlaps between the epitopes were then studied by polyacrylamide gel electrophoresis, competitive binding inhibition, and immunoelectron microscopy. Four groups of epitopes were identified. One group was capable of binding exclusively to the free subunit. The other three groups were identical to those found in immunoelectron microscopy. The data are consistent with the existence of a small number of immunodominant regions on the surface of the Aa6 subunit.

Structure and function of the carboxyl-terminal oxygen-binding domain from the subunit of Octopus dofleini hemocyanin

The C-terminal domain, Od-1, of the 7-domain subunit of Octopus dofleini hemocyanin has been prepared by partial trypsinolysis followed by ion-exchange chromatography. It binds oxygen reversibly and is homogeneous in molecular weight. Its physical properties have been compared with those of the subunit. The domain molecular weight is found by sedimentation equilibrium to be 4.7 X 10(4), in excellent agreement with the result recently obtained in our laboratory from cDNA sequencing of this domain [Lang, W. H. (1988) Biochemistry (preceding paper in this issue)]. It has a sedimentation coefficient of 3.8 S. Both the molecular weight and sedimentation coefficient are consistent with the domain constituting approximately one-seventh of the Mr 3.5 X 10(5) subunit. Its amino acid composition and carbohydrate content differ significantly from that of the whole subunit, confirming the heterogeneity in domains previously established on an immunological basis. Circular dichroism predicts similar secondary structure for the domain and subunit. The domain does not self-associate in the presence of Mg2+ but does bind to the whole molecule in a ratio of approximately 1 domain/subunit. The oxygen affinity of this domain is quite low. It shows intrinsic magnesium and Bohr effects similar to those of the whole molecule but of greatly reduced magnitude.

Mapping of six epitopes on haemocyanin subunit Aa 6 by immunoelectron microscopy

Monoclonal antibodies directed against the haemocyanin of the scorpion Androctonus australis were raised in order to map antigenic determinants (epitopes). The method of mapping employed in this study is molecular immunoelectron microscopy. It consists of a direct electron microscopic observation of antigenic molecules labelled with monoclonal antibodies. The epitopes are then localized in a small region of the external surface of the antigenic molecule whose architecture and quaternary structure are well known. Six monoclonal antibodies have been selected and epitopes have been circumscribed within a small area of one subunit among the 24 subunits composing the whole antigenic molecule.

Intramolecular localization of epitopes within an oligomeric protein by immunoelectron microscopy and image processing

Three epitopes have been localized by immunoelectron microscopy on subunit Aa6 of the 4 x 6-meric hemocyanin of the scorpion Androctonus australis. Soluble immunocomplexes composed of monoclonal antibodies and of native hemocyanin were purified, negatively stained with uranyle acetate by the single-layer technique, and examined under the electron microscope (EM). The molecule images were digitized, aligned, and submitted to correspondence analysis according to the method of Van Heel and Frank (Ultramicroscopy 6:187-194, 1981). A high-precision localization of the attachment point of the Fab arm to the antigen was achieved through a careful analysis of the average images. This method easily allowed the discrimination of epitopes located in different domains (Mr 20 kDa) of the same subunit. Nonoverlapping epitopes located in the same structural domain of subunit Aa6 could be distinguished by the stain exclusion patterns of their Fab arms. The method is general and may be used for epitope mapping in any antigen producing definite EM views.

Approach to the direct intramolecular localization of antigenic determinants in Androctonus australis hemocyanin with monoclonal antibodies by molecular immunoelectron microscopy

Monoclonal antibodies (mAb) directed vs. subunits from hemocyanin (Hc) of the scorpion Androctonus australis were used in molecular immunoelectron microscopy (MIEM) to directly localize the epitopes within the subunits. Four types of mAb were used. First, mAb 6302, an IgG clone highly specific for subunit Aa 2, produced with native hemocyanin long strings composed of hemocyanin molecules in the side view and in the 45 degrees view. At lower concentration, "parachute" and "butterfly" structures composed of two Hc molecules and one monoclonal immunoglobin G (IgG) molecule were obtained. Fab fragments prepared from mAb 6302 bound exactly on the top and bottom edges of the molecule. The second type of mAb (6003), directed vs. subunit Aa 2, produced nice immunocomplexes with the free subunit but nothing with the native oligomer. It is suggested that due to steric hindrance or to conformational changes the epitope is not accessible in the native molecule. The third mAb belonged to the IgM class and apparently bound Hc in the Aa 2 area. However, because of the difficulty of separating the immunocomplexes from the residual mAb and the polymorphism of the IgM molecules, monoclonal IgM are no longer used for MIEM. The last type of mAb (5701) had a high affinity and a high specificity for subunit Aa 6. It produced two types of immunocomplexes with native Hc. The two types differed by a 180 degrees rotation around one of the Fab arms. These complexes, which support recent results of Wrigley et al. [Wrigley, N. G., Brown, E. B., & Skehel, J. J. (1983) J. Mol. Biol. 169, 771-774] and of Roux [Roux, K. H. (1983) Eur. J. Immunol. 14, 459-464], indicate that monoclonal IgG have a high degree of rotational flexibility around the Fab arm. Monoclonal antibody 5701 bound exactly at the corner of the molecule in the area where subunit Aa 6 is known to be located. The MIEM approach of the location of the epitope requires the model of the architecture and of the quaternary structure to be very precise. Thus, recent findings of Gaykema et al. [Gaykema, W. P. J., Hol, J. M., Vereijken, J. M., Soeter, N. M., Bak, H. J., & Beintema, J. J. (1984) Nature (London) 309, 23-29] and of Van Heel et al. [Van Heel, M., Keegstra, W., Schutter, W., & Van Bruggen, E. F. J. (1983) Life Chem. Rep., Suppl. Ser. 1, 69-73] led to a reexamination of previous models.(ABSTRACT TRUNCATED AT 400 WORDS)

Cell-free synthesis of hemocyanin from the scorpion Androctonus australis. Characterization of the translation products by monospecific antisera

Translation of Androctonus australis poly(A)-RNA in vitro led to a number of polypeptides products (8-10) of 70-73 kDa analyzed by two-dimensional gel electrophoresis and identified by immunoprecipitation with an anti-(dissociated hemocyanin) antiserum. The translated hemocyanin polypeptides have the same physico-chemical characteristics as authentic hemocyanin subunits. Subunits Aa 2 and Aa 4 have been identified with monospecific antisera characterized (a) by their capability of reacting with their homologous subunit and (b) by their inability of binding to cross-reacting subunits. Each polypeptide chain is coded by a different messenger without significant post-translational events. Hemocyanin could be detected among the translation products of the poly(A)-RNA isolated from the cuticle under the carapace.

Immunological correlates between multiple isolated subunits of Androctonus australis and Limulus polyphemus hemocyanins: an evolutionary approach

Immunological cross-reactivities between isolated subunits of the scorpion Androctonus australis (Aa) and of the horseshoe crab Limulus polyphemus (Lp) hemocyanins were studied using subunit-specific antibodies prepared through immunoadsorption to pure immobilized subunits. Rocket immunoelectrophoreses of the various subunits of both hemocyanins were carried out at constant antigen concentration against the various subunit-specific antibody preparations. Then the data were analyzed through factorial correspondence analysis and compared to the respective intramolecular locations of the subunits in both hemocyanins. The results show that the dimeric subunits located in the central part of each (4 X 6)meric structure (Aa whole molecule and Lp half molecule) were strongly preserved. In addition, the (8 X 6)mer-forming subunit of Lp hemocyanin (LpIV) and the subunit occupying the same intramolecular position in Aa hemocyanin (Aa5A) were also strongly preserved. Besides the strong antigenic relatedness, less pronounced crossed immunoprecipitations or no precipitation at all were observed between subunits with homologous positions suggesting a minor structural and/or functional roles for these subunits. All the antigen-antibody combinations leading to an absence of immunoprecipitation were screened for the presence of soluble immunocomplexes by radioimmunological tests. In all cases, soluble immunocomplexes were observed. These results suggest the following evolution scenario. First, the central dimeric subunits, responsible of the dodecamer aggregation (Aa3C and 5B and LpV and VI) were already differentiated when Merostomata diverged from Arachnida. Second, the differentiation of the (8 X 6)mer-forming subunit occurred in the Merostomata ramification in a preserved subunit already possessing a functional advantage. Third, the differentiation of subunits Aa3A and Aa3B recently occurred in the scorpion ramification.

Architecture of Limulus polyphemus hemocyanin

The architecture of the 48-meric hemocyanin of the horseshoe crab Limulus polyphemus has been determined from electron micrographs of whole (48-mer) molecules and half- (24-mer) molecules. The assembly of hexamers of kidney-shaped subunits can produce two dodecameric enantiomorphs, designated as right and left. The assembly of 24-mers can again result in two enantiomorphs. By taking into account the rocking effect described by Van Heel and Frank [Van Heel, M., & Frank, J. (1981) Ultramicroscopy 6, 187-194], we deduced that the 24-meric half-molecule is made up of two copies of the left dodecameric enantiomorph. In addition, the two constituent dodecamers of the half-molecule are shifted with respect to a symmetric head-to-tail arrangement, which makes it possible to distinguish two different faces of the 24-mer, termed flip and flop. A model of the whole molecule was built from two copies of the 24-meric half-molecule. This model presents the four distinct views observed in the electron microscope (pentagon, ring, cross, and bowtie). In addition, the model shows the pentagonal view to exist in two varieties: symmetric and asymmetric. An analysis of electron micrographs presenting the pentagonal view by image processing using the statistical technique of correspondence analysis confirmed the existence of two types of pentagonal view, representing projections of a molecule built from two copies of the left 24-meric enantiomorph. In addition, the best fit between the averaged molecule images and the possible models was observed with a flop-flop inter 24-mer contact. The final model is shown in a series of stereo views produced by computer graphical techniques.

A refined quaternary structure of Androctonus australis Hemocyanin

The quaternary structure of the (4 X 6)-mer hemocyanin from the Scorpion Androctonus australis previously published [Lamy, J., Bijlholt, M. M. C., Sizaret, P.-Y., Lamy, J., and van Bruggen, E. F. J. (1981) Biochemistry, 20, 1849-1856] has been refined. The relative positions in the half molecule of subunits Aa 3A and Aa 3B compared to those of Aa 3C and Aa 5B have been established by double labeling of the (2 X 6)-mer with binary mixtures of subunit-specific Fab fragments. The results show that subunits Aa 3B and Aa 5B are located in the same hexamer while Aa 3A and Aa 3C are in the other half of the (2 X 6)-mer. The choice of the enantiomer was deduced from a careful examination of electron micrographs of the native molecule. Finally a position was assigned to each of the 24 subunits on the flip and flop faces as defined by Van Heel and Frank [Ultramicroscopy, 6, 187-194 (1981)].

Hemocyanins in spiders, VII. Immunological comparison of the subunits of Eurypelma californicum hemocyanin

The isolated subunites of Eurypelma californicum hemocyanin were studied by aid of antibodies raised against whole, dissociated hemocyanin. The proportion of impurities was found to be low in almost all subunits. There was no cross reaction between the individual chains, and the total number of antigenically different subunits was found to be seven, confirming results obtained by different methods. If an artificial mixture prepared from purified subunits is compared to whole, dissociated hemocyanin, an overall very similar pattern is obtained but differences appear which are due to specific interaction.--The dimeric subunit 4D was shown to be a heterodimer (asymmetric dimer) composed of chains b and c4.