Search

In the absence of perfect reference standard, classical techniques result in biased diagnostic accuracy and prevalence estimates. By statistically defining the true disease status, latent class models (LCM) constitute a promising alternative. However, LCM is a complex method which relies on parametric assumptions, including usually a conditional independence between tests and might suffer from data sparseness. We carefully applied LCMs to assess new campylobacter infection detection tests for which bacteriological culture is an imperfect reference standard. Five diagnostic tests (culture, polymerase chain reaction and three immunoenzymatic tests) of campylobacter infection were collected in 623 patients from Bordeaux and Lyon Hospitals, France. Their diagnostic accuracy were estimated with standard and extended LCMs with a thorough examination of models goodness-of-fit. The model including a residual dependence specific to the immunoenzymatic tests best complied with LCM assumptions. Asymptotic results of goodness-of-fit statistics were substantially impaired by data sparseness and empirical distributions were preferred. Results confirmed moderate sensitivity of the culture and high performances of immunoenzymatic tests. LCMs can be used to estimate diagnostic tests accuracy in the absence of perfect reference standard. However, their implementation and assessment require specific attention due to data sparseness and limitations of existing software.

Introduction Dans les etudes diagnostiques, les modeles a classes latentes (MCL) visent a estimer de facon non biaisee les performances diagnostiques de nouveaux tests binaires lorsque le test de reference est imparfait et que le niveau d’imperfection est inconnu. Cependant, lorsqu’il s’agit d’evaluer les performances diagnostiques de tests dont le seuil de positivite a ete anterieurement determine en comparaison avec le test de reference imparfait, l’utilisation des MCL permet-elle encore d’obtenir des estimations des performances non biaisees ? L’objectif de ce travail etait d’evaluer, a partir d’une etude de simulations utilisant les MCL, le biais sur les performances diagnostiques de tests binaires dont le seuil de positivite a ete determine en comparaison a un test de reference imparfait. Methode Le biais relatif et le taux de couverture de l’intervalle de confiance a 95 % des sensibilites (Se) et specificites (Sp) des differents tests ont ete evalues selon differents scenarios faisant varier le niveau d’imperfection du test de reference (Se = 0,70/0,85 ou Sp = 0,90/0,95) et des nouveaux tests (Se de 0,75 a 0,90 et Sp de 0,80 a 0,95), la regle de determination du seuil de positivite des nouveaux tests (Se ou Sp minimale de 90 %), la prevalence de la condition (entre 10 % et 50 %) et la taille de l’echantillon (500, 1000 et 5000 individus). Les performances diagnostiques ont ete estimees par un MCL standard. Pour chaque scenario, 500 echantillons de donnees ont ete simules. Resultats Dans le cas d’un test de reference avec une sensibilite imparfaite et des seuils de positivite determines selon une specificite minimale, la sensibilite des nouveaux tests diagnostiques etait sous-estimee (jusqu’a −50 % dans les scenarios etudies les plus biaises) tandis que leur specificite etait surestimee (pour atteindre une Sp proche de 100 % dans les scenarios etudies les plus biaises). Les biais augmentaient avec l’imperfection du test de reference et la prevalence de la condition. Dans le cas d’un test de reference avec une specificite imparfaite et des seuils de positivite determines selon une sensibilite minimale, la specificite des nouveaux tests etait sous-estimee (jusqu’a −75 % dans les scenarios etudies les plus biaises) tandis que leur sensibilite etait surestimee (pour atteindre une Se proche de 100 % dans les scenarios etudies les plus biaises). Les biais augmentaient avec l’imperfection du test de reference et la diminution de la prevalence de la condition. Dans les deux configurations, la taille de l’echantillon n’avait pas d’influence sur le niveau du biais relatif. Les taux de couverture etaient fortement alteres jusqu’a presenter une valeur nulle. Dans les autres configurations, les estimations de la performance diagnostique des nouveaux tests obtenues par les MCL etaient valides. Hormis dans certaines configurations extremes, les performances diagnostiques du test de reference etaient correctement estimees. Conclusion Lorsque le seuil de positivite de nouveaux tests diagnostiques a ete determine en comparaison a un test de reference imparfait, l’utilisation des MCL pour evaluer les performances diagnostiques de ces tests binaires en tenant compte de l’imperfection du test de reference peut mener a des estimations severement biaisees. Des nouvelles strategies de correction doivent donc etre proposees/mises en œuvre au moment de la determination du seuil de positivite de nouveaux tests diagnostiques ou a defaut au moment de l’evaluation de leurs performances diagnostiques.

Objectives Combining noninvasive tests increases diagnostic accuracy for staging liver fibrosis in hepatitis C virus (HCV)-infected patients, but this strategy remains to be validated in HIV/HCV coinfection. We compared the performances of transient elastography (TE), Fibrotest (FT), the aspartate aminotransferase-to-platelet ratio index (APRI) and two algorithms combining TE and FT (Castera) or APRI and FT (SAFE) in HIV/HCV coinfection. Methods One hundred and sixteen HIV/HCV-coinfected patients (64% male; median age 44 years) enrolled in two French multicentre studies (the HEPAVIH cohort and FIBROSTIC) for whom TE, FT and APRI data were available were included in the study. Diagnostic accuracies for significant fibrosis (METAVIR F ≥ 2) and cirrhosis (F4) were evaluated by measuring the area under the receiver-operating characteristic curve (AUROC) and calculating percentages of correctly classified (CC) patients, taking liver biopsy as a reference. Results For F ≥ 2, both TE and FT (AUROC = 0.87 and 0.85, respectively) had a better diagnostic performance than APRI (AUROC = 0.71; P

Chlamydia trachomatis, Neisseria gonorrhoeae, and Mycoplasma genitalium screening during pregnancy is not performed routinely in France. We conducted the first prospective study in 1004 women attending for routine antenatal care to determine the prevalence and risk factors for these bacterial infections. The overall prevalence of C. trachomatis, N. gonorrhoeae, and M. genitalium infections was 2.5%, 0%, and 0.8%, respectively. In patients aged 18-24 years, the prevalence increased to 7.9% for C. trachomatis and to 2.4% for M. genitalium. C. trachomatis infection was associated with age

Objectif de l’etude Comparaison de la tumorectomie et l’exerese totale du mesorectum chez les patients en situation de bonne reponse a une chimioradiotherapie pour un cancer du rectum. Materiel et methode Les patients atteints de tumeur de stade T2 ou T3 (inferieure ou egale a 4 cm) ont recu une chimioradiotherapie (Cap50, capecitabine). Les patients en situation de bonne reponse clinique (tumeur residuelle ≤ 2 cm) ont ete randomises entre une tumorectomie et l’exerese totale du mesorectum. Le groupe qui a eu une tumorectomie, en cas de reponse ypT2-3 ou R1, a eu une exerese totale du mesorectum complementaire. L’objectif principal etait de montrer la superiorite de la tumorectomie sur la l’exerese totale du mesorectum par un critere composite incluant deces, recidive, morbidite et effets secondaires a 2 ans. Les objectifs secondaires incluaient la reponse tumorale, une survie sans recidive et globale a 3 ans. Resultats De mars 2007 a septembre 2012, 148 patientes en situation de bonne reponse (sur 186 patients pris en charge par chimioradiotherapie) ont ete randomises et les dossiers de 145 patients ont ete analyses : 74 dans le groupe qui a eu une tumorectomie et 71 dans celui qui a eu une exerese totale du mesorectum. Dans le groupe qui a eu une tumorectomie, 26 patients ont eu une l’exerese totale du mesorectum complementaire. En intention de traiter, il n’y avait pas de difference entre les deux groupes pour tous les composants du critere composite. Cependant, les analyses complementaires sur la morbidite chirurgicale (12 % contre 22 % contre 46 %, p = 0,003) et les effets secondaires (19 % contre 30 % contre 59 %, p = 0,001) ont montre l’influence du type de chirurgie, tumorectomie contre l’exerese totale du mesorectum contre l’exerese totale du mesorectum complementaire, respectivement. L’analyse des resultats histopathologiques a montre 61 % de bonnes reponses (ypT0-1) ainsi qu’une association entre les reponses ganglionnaire et tumorale avec 0, 0, 8 % et 40 % de tumeurs ypN1 pour respectivement les tumeurs ypT0, T1, T2 et T3. A 3 ans, il n’y avait pas de difference significative de taux de recidive locale (5,4 % contre 5,6 %) ni de probabilite de survie sans maladie (78,3 % contre 76,1 %). Conclusion L’etude n’a pas montre la superiorite de la tumorectomie par comparaison a la l’exerese totale du mesorectum, en raison d’un taux eleve d’exerese totale du mesorectum complementaires, responsable d’une augmentation de la morbidite et des effets secondaires dans le groupe qui a eu une tumorectomie. Une meilleure selection des patients, en evitant les exereses totales du mesorectum non necessaires, pourrait ameliorer la strategie.

Combining noninvasive tests increases diagnostic accuracy for staging liver fibrosis in hepatitis C virus (HCV)-infected patients, but this strategy remains to be validated in HIV/HCV coinfection. We compared the performances of transient elastography (TE), Fibrotest (FT), the aspartate aminotransferase-to-platelet ratio index (APRI) and two algorithms combining TE and FT (Castera) or APRI and FT (SAFE) in HIV/HCV coinfection.One hundred and sixteen HIV/HCV-coinfected patients (64% male; median age 44 years) enrolled in two French multicentre studies (the HEPAVIH cohort and FIBROSTIC) for whom TE, FT and APRI data were available were included in the study. Diagnostic accuracies for significant fibrosis (METAVIR F ≥ 2) and cirrhosis (F4) were evaluated by measuring the area under the receiver-operating characteristic curve (AUROC) and calculating percentages of correctly classified (CC) patients, taking liver biopsy as a reference.For F ≥ 2, both TE and FT (AUROC = 0.87 and 0.85, respectively) had a better diagnostic performance than APRI (AUROC = 0.71; P0.005). Although the percentage of CC patients was significantly higher with Castera's algorithm than with SAFE (61.2% vs. 31.9%, respectively; P0.0001), this percentage was lower than that for TE (80.2%; P0.0001) or FT (73.3%; P0.0001) taken separately. For F4, TE (AUROC = 0.92) had a better performance than FT (AUROC = 0.78; P = 0.005) or APRI (AUROC = 0.73; P = 0.025). Although the percentage of CC patients was significantly higher with the SAFE algorithm than with Castera's (76.7% vs. 68.1%, respectively; P0.050), it was still lower than that for TE (85.3%; P0.033).In HIV/HCV-coinfected patients, TE and FT have a similar diagnostic accuracy for significant fibrosis, whereas for cirrhosis TE has the best accuracy. The use of the SAFE and Castera algorithms does not seem to improve diagnostic performance.

The main source of key medical information consists in original articles published in peer-reviewed biomedical journals. Reported studies use increasingly sophisticated statistical and epidemiological approaches that first require a solid understanding of core methods. However, such understanding is not widely shared among physicians. Our aim was to assess whether the basic statistical and epidemiological methods used in original articles published in general biomedical journals are taught during the first years of the medical curriculum in France.We selected original articles published in The New England Journal of Medicine, The Lancet, and The Journal of the American Medical Association, over a period of six months in 2007 and in 2008. A standardized statistical content checklist was used to extract the necessary information in the "Abstract", "Methods", "Results", footnotes of tables, and legends of figures. The methods used in the selected articles were compared to the national program and the public health program of biostatistics and epidemiology taught during the first six years of medical school.The 237 analyzed original articles all used at least one statistical or epidemiological method. Descriptive statistics, confidence interval and Chi(2) or Fisher tests, methods used in more than 50% of articles, were repeatedly taught throughout the medicine curriculum. Measures of association, sample size, fit and Kaplan-Meier method, used in 40 to 50% of articles, were specifically taught during training sessions on critical reading methods. Cox model (41% of articles) and logistic regression (24% of articles) were never taught. The most widely used illustrations, contingency tables (92%) and flowcharts (48%), were not included in the national program.More teaching of the core methods underlying the understanding of sophisticated methods and illustrations should be included in the early medical curriculum so that physicians can read the scientific literature critically for their medical education.

For a long time, amino acid or peptide containing lipids were considered as unusual components of bacteria. Isolation was mainly confined to “lipo-amino acids” which appeared to be transitory intermediates in cellular metabolism. Not much progress has been made on their metabolism.

Summary Background The pathogenesis of Crohn’s disease (CD) involved microbial factors. Some Helicobacter species, the so-called entero-hepatic Helicobacters (EHH), can naturally colonize the intestinal surface and have been detected in humans. Aim To look for an association between CD and the presence of EHH DNA in intestinal biopsies. Methods Two groups of patients were included prospectively in a multicentre cross-sectional study: CD patients with an endoscopic post-operative recurrence within 2 years following a surgical resection and controls screened for colorectal polyps or cancer. Intestinal biopsies were taken for Helicobacter culture and Helicobacter 16S DNA detection. If positive, the EHH species were identified with specific PCRs, sequencing and denaturing gradient gel electrophoresis. Results In the 165 included patients (73 CD and 92 controls), Helicobacter cultures were negative. PCR was positive in 44% of CD and 47% of controls. After age-adjustment, CD was significantly associated with EHH in intestinal biopsies (OR = 2.58; 95%CI: 1.04–6.67). All EHH species detected were identified as Helicobacter pullorum and the closely related species Helicobacter canadensis. Conclusion Crohn’s disease is associated with the presence of EHH species DNA in intestinal biopsies after adjustment for age. Whether these species play a role in the pathophysiology of CD remains to be determined.

Chemotaxonomic identification of coryneform bacteria metabolizing b-caryophyllene was attempted. The following phospholipids were identified as main components of the bacterial extracts: cardiolipids, phosphatidylethanolamine, phosphatidylinositol and mannosides of phosphatidylinositol. Saponification of the lipid extracts gave a mixture of hydroxylated and nonhydroxylated fatty acids. Among the latter, oleic and tuberculostearic acids were identified. The hydroxylated fatty acids were analysed by thin-layer chromatography and mass spectrometry (as methyl esters). From the results thus obtained, the strains appeared to be more closely related to the genus Rhodococcus than to the genus Nocardia.

When the sensitivity of an assay used to quantify a marker is poor, some of the values are below the detection limit resulting in left-censoring. Analysis of such data requires appropriate statistical techniques. In this study, we aimed at comparing various methods used to deal with left-censored outcome in regression analysis.The application was a real study evaluating the performance of procalcitonin for the diagnosis of bacterial infections among elderly patients. Among 85 patients, eleven had a procalcitonin value below the detection limit. A simulation study was then performed with data sampled according to a Gaussian distribution with parameters estimated on observed data. Various levels of left-censoring were simulated (13, 25 and 50%). A linear regression model was used to explain procalcitonin variations according to another marker, C reactive protein. To handle left-censoring, several methods were used: complete case analysis, simple imputation and multiple imputation methods, and parametric modelling. In the simulation study, estimations according to different methods were compared in terms of bias and mean square error according to each left-censoring level. Estimations obtained with real data were also compared according to the methods used. All analyses were implemented using SAS software.In the simulation study, parametric modelling using maximum likelihood showed best performances whatever the level of censoring. On the other hand, methods using complete cases and simple imputation by the detection limit were highly skewed. On observed data, estimations of the slope varied slightly according to the methods. However the p-values (Wald test) of beta=0 varied from 0.0001 to 0.13 leading to different decisions according to the method used.Left-censoring handling in data analysis requires special attention, as different methods may yield results leading to different conclusions.

an eyeball phantom that uses a programmable x-y motion stage with a connected joystick to drive the simulated eyeball’s rotational movement. Results: Both tumor (size and location inside the eyeball) and eyeball (size) were fabricated in the size similar to human’s. Eyeball’s movement can be programmed in various speeds and patterns. Experimental data from phantom validation under the CT simulator and from the 2D/ 3D transformation based on the video camera tracking were obtained. For eight testing positions (center of the simulated tumor) in the validation experiment, we compared the ground true coordinate from the CT space and predicted coordinate from our derived transformation. The maximum root-sum-squared error (RSSE) of three coordinates (X, Y, Z) is less than 0.70 mm and the mean-squared error is 0.44 mm among eight test positions. Conclusions: This video camera tracking technique and the 2D/3D transformation are feasible for robotic radiosurgery on ocular tumor. Error analysis indicates that the accuracy is clinically acceptable. Author Disclosure: W. Zhao: None. H. Chen: None. X. Wu: None. G.F. Hatoum: None.

CO08-007–FR Communication dans la vie quotidienne des personnes aphasiques apres accident vasculaire cerebral T. Lagadec , D. Zongo , J. Asselineau , E. Douce , J. Trias , M.F. Delair , M. Michot , B. Darrigrand , J.-M. Mazaux g,∗ a EA 4136, universite Bordeaux-Segalen, hopital IA–Robert-Picque, Bordeaux, France b ISPED, universite Bordeaux-Segalen, Bordeaux, France c Hopital Nouvielle, Mont-de-Marsan, France d Hopital Robert-Boulin, Libourne, France e Centre hospitalier universitaire de Bordeaux, Bordeaux, France f EA 4136, universite Bordeaux-Segalen, hopital Robert-Boulin, Libourne, Bordeaux et Libourne, France g EA 4136, universite Bordeaux-Segalen, CHU de Bordeaux, Bordeaux, France Auteur correspondant.

Objectif Fibrostic est une etude multicentrique nationale dont l’objectif est de determiner les seuils d’elasticite et la performance diagnostique du Fibroscan® dans la prediction de (a) une fibrose Metavir significative ≥ F2, et (b) la cirrhose, sur une grande cohorte de patients atteints d’hepatite chronique virale. Patients et Methodes En 2 ans, 23 centres ont inclus 1 993 patients avec PBH, tests biologiques de fibrose et Fibroscan® simultanes. En accord avec le plan d’analyse defini a priori dans le protocole, le seuil d’elasticite hepatique pour l’analyse principale a ete choisi pour maximiser la sensibilite (Se) sous contrainte d’une specificite (Sp) d’au moins 90 %. Il a ete determine par des courbes ROC pour chaque sous-groupe d’infection virale. Resultats L’analyse a porte sur 1 839 patients : 532 (27 %) ne satisfaisaient pas aux pre-requis du test et 1 307 etaient eligibles pour l’analyse principale, d’âge median 47 ans, sexe M/F 69/31 %, le diagnostic clinique de cirrhose etait porte dans 154 cas, compliquee chez 45 patients. La valeur mediane des ALAT etait de 1,6N (m = 2,4+/−3,1). Les valeurs moyennes d’albuminemie, de bilirubinemie, TP et numeration des plaquettes etaient normales. Virologie - Infection par le VHC : 913 patients (dont 18 VHC+VHB), ARN VHC positif 91 %, genotype 1 65 % ; infection par le VHB : 284 patients ; co-infection VIH+VHB ou VHC : 100 patients. La taille moyenne des PBH etait de 25 ± 11 mm avec 17,5 ± 10 espaces portes. En cas d’infection VHC, le seuil diagnostique de fibrose ≥ F2 (prevalence 62 %) etait de 10,3 kPa, (Se 39 %, Sp 91 %, VPP 87 %, VPN 48 %) avec une AUC ROC de 0,74. Le seuil diagnostique de prediction de la cirrhose VHC (prevalence 14 %) etait de 13,4 kPa, (Se 71 %, Sp 90 %, VPP 54 %, VPN 95 %), avec une AUC ROC de 0,89. Conclusion Le Fibroscan® montre une bonne performance pour ecarter le diagnostic de cirrhose virale mais pas pour predire une fibrose significative. Le seuil d’elasticite le plus adapte ne pourra etre choisi qu’en fonction de la place de cet examen dans la strategie diagnostique de la fibrose hepatique. Remerciements, financements, autres Financement DHOS, Gestion : F Tubach, Y Costa, V Olympio, M Achahboun (URC Bichat APHP).

For a long time, amino acid or peptide containing lipids were considered as unusual components of bacteria. Isolation was mainly confined to “lipo-amino acids” which appeared to be transitory intermediates in cellular metabolism. Not much progress has been made on their metabolism.

Chemotaxonomic identification of coryneform bacteria metabolizing b-caryophyllene was attempted. The following phospholipids were identified as main components of the bacterial extracts: cardiolipids, phosphatidylethanolamine, phosphatidylinositol and mannosides of phosphatidylinositol. Saponification of the lipid extracts gave a mixture of hydroxylated and nonhydroxylated fatty acids. Among the latter, oleic and tuberculostearic acids were identified. The hydroxylated fatty acids were analysed by thin-layer chromatography and mass spectrometry (as methyl esters). From the results thus obtained, the strains appeared to be more closely related to the genus Rhodococcus than to the genus Nocardia.

To increase the application of lipid analysis for taxonomic purposes, a standardized method would be useful. This method has to be simple enough to be used routinely. Such a method, based on thin-layer and gas chromatography, is proposed. It can be applied to any kind of bacteria when a few milligrams of cells are available. Comparison of the results provided by this analysis with data found in the literature (given in a schematic form in six tables) allows the identification of the studied strain in the most favorable cases, or gives information for the choice of conclusive complementary tests in the other cases.

Summary When cultivated on solid medium or in stationary liquid medium, most of the 11 studied strains of Bacillus benzoevorans grew as unbranched, flexible, immotile filaments (or trichomes) of undefined length and 1.8 μm diameter. They were enclosed in a sheath giving an overall diameter of 3.6 μm. When cultivated in vigorously shaken liquid medium, several strains grew as separate rods (1.8×3–16 μm) which had rounded ends and were motile by means of peritrichous flagella. This pleomorphism was due to an adaptation and not to mutation. The endospores were oval (1.8×2.2 μm), did not deform the filament and contained dipicolinic acid. The cytoplasm had a granular aspect due to the presence of poly-β-hydroxybutyric acid. The Gram reaction was variable but the Gram type was positive. Isolated and purified sheath contained 73% proteins, 12.9% reducing sugars and 8% lipids. In stationary liquid medium, the culture had a mycelial aspect and a thick pellicle formed at the surface. Colonies were circular, flat, opaque, whitish, mat and compact; they had irregular edges, spread out on the surface and did not adhere to the agar.

On the basis of their biological properties (cell morphology, staining reaction, cellular inclusions, guanine plus cytosine content of deoxyribonucleic acid, genome size, degree of deoxyribonucleic acid homology with Corynebacterium diphtheriae [Approved Lists, 1980] strain PW8, pattern of enzymic capacities) and chemical structures (peptidoglycan type, major cell wall polysaccharide, characteristic fatty acids, major phospholipids and glycolipids), two brevibacteria are transferred to the genus Corynebacterium. Brevibacterium vitarumen (Bechdel, Honeywell, Dutcher, and Knutsen) Breed (Approved Lists, 1980) is transferred as Corynebacterium vitarumen (Bechdel, Honeywell, Dutcher, and Knutsen) comb. nov., and Brevibacterium liquefaciens Okabayashi and Masuo (Approved Lists, 1980) is transferred as Corynebacterium liquefaciens (Okabayashi and Masuo) comb. nov.

On the basis of its biological properties (cell morphology, staining reactions, cellular inclusions, guanine plus cytosine content of deoxyribonucleic acid, degree of deoxyribonucleic acid homology with Corynebacterium diphtheriae strain PW8, and pattern of enzymic capacities) and chemical structures (peptidoglycan type, major cell wall polysaccharide, major phospholipids and glycolipid, as well as characteristic cell wall fatty acids), Microbacterium flavum Orla-Jensen (synonym: Mycobacterium flavum Jensen) is renamed Corynebacterium flavescens nom. nov. The specific epithet “flavum” has been changed to “flavescens” in order to avoid the creation of a later homonym of Corynebacterium flavum Kisskalt and Berend 1918. Strain 8 of Orla-Jensen (= ATCC 10340 = NCIB 8707) is herein designated the type strain of C. flavescens. An updated description of the organism is given.

6,6'-Dicorynomycoloyl esters of alpha,alpha'- or beta;beta'-trehalose were synthesized and tested on isolated rat liver mitochondria. In contrast to the well known site-II-specific uncoupling effect of the alpha,alpha'-trehalose derivative, the beta,beta'-trehalose derivative exhibited only non-specific inhibition of active respiration in the presence of glutamate. It is proposed that this unexpected difference between the two isomers could arise from conformational differences in the carbohydrate moiety of the glycolipids, which would render the beta,beta' isomer unable to recognize targets specific to the alpha,alpha'-trehalose glycolipid.

Summary The eleven strains studied were prototrophic and did not grow in media containing only 1% Bacto-peptone or Bacto-tryptone; they grew rapidly in media containing 0.4% yeast extract and 0.2% sodium acetate or benzoate. The maximal growth temperature ranged from 39 to 45°C. Six aliphatic acids, four aromatic acids and five phenols were used as sole carbon and energy sources by the 11 strains. Carbohydrates and amino acids (except for glycine)were not used as carbon and energy sources. Nitrate (but not nitrite) was used anaerobically as a respiratory electron acceptor. Nitrous oxide was used and reduced to N2 by only 3 strains. The mean guanine-plus-cytosine content of the DNA was 41.3±1.1 mol%. Morphologically and nutritionally, the bacteria described are clearly different from the 5 known species of the first morphological group whose cells have a diameter greater than 1 μm: Bacillus megaterium, B. cereus, B. cereus var. mycoides, B. macroides, B. badius, and B. fastidiosus. Strain B1 (=CCM 3364) is the holotype of Bacillus benzoevorans sp. nov.

Summary Analysis of the lipids of seven species of Micromonospora showed the constant presence of five phospholipids and three glycolipids. The phospholipids were identified as cardiolipids, phosphatidylethanolamine, phosphatidyl-inositol, mannosides of phosphatidyl-inositol, and possibly lysocardiolipids. The glycolipids were monoglucosyl-diglycerides, diglucosyldiglycerides (with a small amount of galactosyl-glucosyl-diglycerides) and esters of fatty acids and trehalose. The simultaneous presence of glucosyldiglycerides and trehalose esters might be specific of the genus Micromonospora . In the seven species studied, a large amount of the total fatty acids(obtained by direct saponification of the cells) consisted of iso and anteiso fatty acids. Three groups could be distinguished, having respectively iso -C 15 , iso -C 16 and n -C 17 fatty acids as the main members. A monounsaturated C 17 acid has been identified as hepatadec-9,10-enoic acid.

The composition of the Pseudomonas solanacearum lipolysaccharide (LPS) was found to be similar to that described for the LPS of enterobacteria. The lipid A contained fatty acids and glucosamine in a molar ratio of 5:2. The LPS fraction contained 2-keto-3-deoxyoctulosonic acid, L-glycero-D-mannoheptose, hexoses (glucose, rhamnose, and glucosamine), and a pentose (xylose). The LPSs from the wild-type strain (GMI1000), from the spontaneous rough mutant (GMI2000), and from their respective acridine orange-resistant (Acrr) mutants (GMI1178 and GMI2179) contained the same component sugars in their polysaccharide moieties, but the relative amounts of each sugar varied greatly. Spontaneous mutation to the rough type was characterized by a decrease in the ratio of rhamnose to glucose, whereas a reverse effect was seen for the acridine orange resistance mutation from the parent strains (GMI1000 and GMI2000) to the respective mutant strains (GMI1178 and GMI2179). The exopolysaccharide (EPS) from GMI1000 was found to be composed of two fractions: a heteropolysaccharide (galactosamine, glucose, and rhamnose) excluded from Sephadex G-50 and an additional glucan with a lower molecular weight. Strains GMI1000 and GMI1178 produced comparable amounts of EPS, GMI2179 synthesized less EPS, and GMI2000 produced no detectable EPS. High-pressure liquid chromatography and 13C nuclear magnetic resonance analyses revealed some differences between these EPSs. The glucan fraction seemed to be the major component of the EPS from GMI2179, whereas GMI1000 and GMI1178 EPSs contained both fractions and appeared to differ in the structures of their heteropolysaccharide fractions. Viscosity measurements confirmed differences between whole EPSs produced by the three strains.

Summary Twenty encapsulated strains of Flavobacterium multivorum were isolated from soil by elective culture in minimal medium containing inulin as the sole source of carbon and energy. These strains were compared with the type strain and with 5 other strains (including 4 clinical strains) of F. multivorum . Of 168 substrates tested, 18 carbohydrates were used as the sole carbon and energy source by all 26 strains. A few amino acids were used by some strains. The yellow pigment produced was found to be a carotene and its production was photo-inducible. The presence of a cytochrome c oxidase of the aa 3 type was suggested by absorption spectra. The major cell lipids were phosphatidylethanolamine and sphingolipids; the major fatty acids were 13-methyltetradecanoic and 2-hydroxy-13-methyltetradecanoic acids. Soil and clinical strains of F. multivorum showed roughly similar patterns of antibiotic multiresistance. The average G+C content of the DNA of 11 strains was 40.8±1.5 mol %.

Summary Thirty strains were isolated from pasteurized soil samples by enrichment culture in aerobiosis at 32°C in a minimal medium containing one of the following compounds as sole source of carbon and energy: quinate, p-hydroxybenzoate, phthalate, isophthalate or trimellitate. These bacteria were rods (0.8×2–7 μm), motile by peritrichous flagella. Endospores were oval (1.4–1.8×2 μm) and distinctly swelled the sporangia. The Gram reaction was variable but the Gram type was positive. Colonies were smaller on peptone (0.4%) agar than on minimal salts-glucose (0.2%) agar. The following characters were always present: growth in the presence of lysozyme, cytochrome c oxidase, catalase, nitrate assimilation, urease, amylase and L-glutamate dehydrogenase. The cells contained glycogen. In anaerobiosis, glucose was not fermented and nitrate was not used as a respiratory acceptor of electrons. Of 215 substrates tested, 31 (including 9 aromatic compounds) were used as sole carbon and energy sources by all 30 strains, and 38 substrates (including 13 aromatic compounds) were used by only some of them; 146 substrates (including 49 aromatic compounds) were not used by any of the 30 strains. No amino acid could be used as sole carbon and energy source. Numerical analysis of the 30 strains showed an aggregate cluster made of 5 phena. The mean G+C content of the DNA was 55±0.6 mol % The described bacteria are clearly different from the 2 known species of the second morphological group which cannot ferment carbohydrates: Bacillus brevis and B. azotoformans. Strain Q1 (ATCC 29948) is the holotype of Bacillus gordonae sp. nov.

The free lipids of Flavobacterium meningosepticum were separated by thin layer chromatography, and the main lipid fractions were analysed by FAB (fast atom bombardment) mass spectrometry. The major products were di-iso-C 15 - and iso-C 15 -iso-C 17 -phosphatidylethanolamine, and two ninhydrine + and phosphorus − fractions. The structures of the latter two fractions were established as ornithine lipids by using MIKE (mass ions kinetic energy) mass spectrometry, GC/MS (gas chromatography coupled with mass spectrometry) and conventional methods. The presence of small amounts of sphingolipids with C 17 - and C 16 -sphinganines was demonstrated. F. meningosepticum can be distinguished from F. multivorum and F. spiritivorum by easy characterization of the ornithine lipids by thin layer chromatography.

Summary Four strains of Bacillus circulans sensu stricto (ATCC 4513 (typestrain), 4515, 4516, and 4530) were subjected to 236 morphological, biochemical and physiological tests, including 162 carbon source utilization tests. B. circulans s. s. is a species of facultative anaerobes able to ferment carbohydrates. Many carbohydrates and a few aliphatic acids were used as sole carbon and energy sources, but amino acids were not. The guanine-plus-cytosine content of DNA (four strains studied) was 37.2±0.4 mol % (mean±standard deviation).

Resume Le dosage de l'alanine, de la valine et de la leucine peut se faire par transformation de chacun des acides amines en les aldehydes correspondants, sous l'action de l'hypochlorite de sodium. L'aldehyde forme est entraine par la vapeur d'eau ou par un courant d'air, et recueilli dans du bisulfite. Le dosage de l'alanine se ramene ainsi au dosage de l'acetaldehyde; celui-ci est specifique si l'on a eu soin d'eliminer prealablement l'acide aspartique du milieu reactionnel, et si on dose l'aldehyde forme soit par le nitroprussiate et la piperazine, soit par le p-hydroxydiphenyle. La separation des aldehydes correspondant a la valine et a la leucine se fait par entrainement et distillation fractionnee, en presence de tetrachlorure de carbone, dans une colonne suffisamment puissante. Dans ces conditions, il est possible de doser dans un melange d'acides amines quelconques, et separement, l'alanine, la valine et l'ensemble des leucines. La precision des resultats obtenus est de l'ordre de 5%.

The three main components of the mixture of corynomycolic acids (α-branched, β-hydroxy acids of the general formula R1—CHOH—CHR2—COOH) isolated from a strain of Corynebacterium hofmanii are: C32H64O3 (corynomycolic acid); C34H66O3 (corynomycolenic acids); C36H68O3 (corynomycoldienic acid). By thin-layer chromatography on silver nitrate-impregnated Silicagel G, we were able to separate the hydroxy acids having 0, 1 or 2 double bonds. The pyrolytic chromatography of their methyl esters indicates that the mono-unsaturated acids form a mixture of C34-acids containing the double bond either in the main chain or in the side chain; the di-unsaturated acid fraction is constituted solely by a C36-acid with a double bond in each fragment arising from pyrolysis. By oxidative cleavage, we find that the position of the double bond is 9, 10 (as in oleic acid). According to the nature of the main chain (aldehyde fragment of pyrolysis) and of the side chains (ester fragment of pyrolysis), some suggestions are made about the selective use of fatty acids in the biosynthesis of corynomycolic acids.

Several glycolipid fractions have been isolated from the bound lipids of N. brasiliensis. The main fraction was characterized as a monoester of nocardomycolic acids and D-arabinose. Fractionation of the free lipids yielded a glycolipid fraction consisting of nocardomycouc acids, arabinose, galactose and several amino acids. This last fraction appears to be similar to wax D of Mycobacteria; mild acid treatment of this fraction produced arabinose mononorcardomycolate.

The main fatty acids of the lipids of Brucella abortus (strain Scherle II, or 1119) and of Brucella melitensis (strain M-15) are palmitic acid, a C18 monoenoic acid an a C19 cyclopropane fatty acid. By chemical degradation, it is shown that the unsaturated C18 acid is cis-vaccenic acid. The C19 cyclopropane acid has been identified with lactobacillic acid, either by permanganate oxidation according to Murray, or by mass spectrometry of the keto-esters obtained by chromic acid oxidation.

The chemical structure of cord factor of Mycobacterium tuberculosis has been established as trehalose-6,6'-dimycolate (I), C186H366O17±10CH2. Alkaline hydrolysis of cord factor yields two molecules of mycolic acid and one molecule of the disaccharide trehalose. Methylation studies show that the two mycolic acid molecules are attached through an ester linkage to the primary hydroxyl groups of trehalose. Identical results were obtained with cord factor from the human strain Brevannes, and the BCG strain.

Resume Le “cord factor” d'une souche H-37 Rv streptomycino-resistante, purifie par chromatographie est une cire presqu'incolore, F 42–44°, [a]J = +30° et presente une composition centesimale en accord avec la formule brute C95H189O9N ± CH2 ± O. La saponification alcaline du “cord factor” libere une molecule d'acide mycolique, C87H174O4 ± 5 CH2. La presence d'un cinstituant osidique est en outre undiquee par des reactions colorees. Ces resultats sont en accord avec ceux de Noll et Bloch 4 qui ont identifie une molecule d'acide mycolique et une molecule de D -glucose dans le “cord factor” d'une autre souche humaine. Un troisieme constituant de faible taille moleculaire et contenant in azote tertiaire n'a pas encore ete identifie.

We isolated a mixture of glyco-peptidolipids (mycosides C) from a non-photochromogenic atypical strain of Mycobacteria. The main component, mycoside C1217, contains a 3 D-hydroxyacid with 28 carbons, two 6-deoxysugars and four amino compounds. The two sugars are 6-deoxytalose and trimethylrhamnose (a minor component contains 3,4-dimethylrhamnose instead of trimethylrhamnose). The four amino compounds are three aminoacids, phenylalanine, allothreonine and alanine, plus an aminoalcool identified with alaninol by gas chromatography of the N-trifluoracetyl derivative. The phenylalanine is N-terminal in the peptide, and linked to the fatty acid by an amide bond. We established, by β-elimination in alkaline medium, that the deoxytalose is linked to the hydroxyl of the allo-threonine by a glycosidic bond. After a mild hydrolysis of the mycoside, we isolated a glycopeptide containing allo-threonine, alanine, alaninol and trimethylrhamnose. By use of the dimethylamino-naphtalene sulfonyl method end of Edman degradation, we have shown that the sequences of this glycopeptide is: allo-threonyl-alanyl-alaninol-trimethylrhamnose All these results, along with those obtained from some other reactions, support the following structure for the mycoside C1217, which is in agreement with mass spectrometric results

The lipids of Corynebacterium ovis (strain Bl/50) were successively extracted by petroleumether, ethanol-ether (1:1, v/v) and chloroform. The bulk of the petroleum-ether extract consists mainly of sodium corynomycolate together with sodium salts of C14, C16 and C18 fatty acids. Some triglycerides and free fatty acids were also detected. Isolated from the ethanol-ether extract were: hydrocarbons, palmitone, free corynomycolic acid, a small amount of free corynomycolenic acid and a peptidolipid fraction. The phospholipid fraction contains a mixture of closely related compounds, having about 3.5% P. By hydrolysis, they are split into lipid components (50–70% of the initial products; mixtures of C16 and C18 fatty acids and corynomycolic acid), and water soluble components (glycerol, glycerophosphoric acid, inositol, arabinose, and small amounts of amino acids). Complex substances, soluble in ether, having a high content of phosphorus (3.2 to 4.5%) were isolated from the chloroform extract. All of them contain glycerol, along with ethanolamine and sugars (glucose, galactose, arabinose, and a fast moving unidentified sugar) in one case, and with amino acids (chemically bound) in two other cases. Moreover, one third of the chloroform extract is made of a fraction (10.8% N and 1.3% P) insoluble in ether, which gives, by hydrolysis, 19% of lipid components and a polypeptide containing all the usual amino acids; the structure of this last fraction is discussed.

Resume 1. 1. Nous decrivons la purification du complexe lipo-polysaccharidique extrait du Bacille Tuberculeux par N. Choucroun et dont les proprietes antigeniques ont ete decrites anterieurement2. 2. 2. Par saponification, une des fractions principales obtenues, facilement soluble dans l'ether de petrole meme a froid, donne 45% d'acide mycolique, II% d'acides gras et 47% d'une partie hydrosoluble constituee essentiellement par un polysaccharide et contenant tout le phosphore et tout l'azote du lipo-polysaccharide. 3. 3. Un quart environ de l'azote est present sous forme d'ammoniac; il y a en outre trois acides amines, dont deux ont ete identifies par chromatographie sur papier; ce sont l'acide glutamique et l'alanine. Le troisieme acide amine est identique a la substance No. 17 que E. Work a recemment isolee des proteines du Bacille Diphterique. Cette substance est neutre, aliphatique et exempte de phosphore et de soufre. Elle ne semble pas se trouver dans les proteines du Bacille Tuberculeux, ni dans la tuberculine. 4. 4. Les trois acides amines semblent engages dans des liaisons peptidiques, un seul groupe -NH2 sur trois etant libre.

The name Bacillus azotoformans (ex Pichinoty, de Barjac, Mandel, Greenway, and Garcia 1976) is revived. Strains of this species are motile and produce terminal or subterminal oval endospores in swollen sporangia. Nitrate, nitrite, and nitrous oxide are denitrified with production of nitrogen. The guanine-plus-cytosine content of the deoxyribonucleic acid of this organism is 39.8 ± 1.2 mol% (mean ± standard deviation). The type strain is ATCC 29788 (= CCM 2849 = CIP R925).