Letters Eosinophil and Mast Cell Staining in Sheep Infected by Oestrus avis Larvae Oestrus ovls is the agent of worldwide naso sinusal myiasis o f sheep and goats. To study the allergic reactions induced by this parasite, at a mucosa] surface, we have hitherto used May-Grunwald Giemsa staining for eosinophils (formol fixative), alcian blue for the mucous mast cells apd Toluidine blue for serous mast cells (both in Camoy fixative). So we have to process three samples: one for each cell type. We applied the technique of Kermanizadeh et al.I to the mucosa of sinus, tur~inates and nasal septum of 2 I sheep naturally infected with Oestrus avis larvae (Linn~ 1761 ), At the same time, we processed similar samples in the normal way. For each sample, reactive cells were counted on 30 microscopy fields (objective x 50) (I0 in subepithelial chorion, I0 in interglandular chodon and I0 in submucosa) (see Table I). Globally, the total numbers of cells are less numerous using Kermanizadeh's method than using separate fixation and staining methods to differentiate the cell types. Nevertheless, the determination coefficient (R2) calculated for mast cells stained with MT fixative ~or by the usual technique was 0.7, and for eosinophils 0.8. These R 2 values indicate a clear relationship between the staining techniques. Our results ccnfirm those of Winter et aL2 However, we did not modify the initial protocc', as they had done. and we didn't observe precipitation, as they did in gut tissue. The results demonstrate, once again, the value of this 'one-step' staining method for eosinophils and mastocyLes. Although it does not permit distinction of serous and mucous mast cells, it does gain in the saving of time. ileferences I Kermanizadeh.P.. Hagan, P. and Crompton. D.W.T. (1995) Parasitology Today I I, 194-196 2 Winter. M. et al. (1995) Parasitology Today I I, 426 Philippe Oorchles Nguyen Van Khanh Paul Cabanie Didier Concordet Ecole l',lationaleV~tdrinaire de Toulouse 23 Chemin des Capelles 31076 Toulouse Cedex, France Table I. Arithmetic mean results for the 63 counts a of each cell type b in each anatomic location c Eosinophils Mast cells MGG J MT ~ Toluidine blue Alcian blue Total Arithmetic mean 92.8 77.7 47. I 63.9 I I 1.5 SO 121.2 71.8 15.4 26.9 38.5 Median 99 78 45 64.8 I 13.3 MT • 96.7 35.6 98.S ale. 9450 microscopic fields. cSubepithe~ial chorion, interglandular chorion, submucosa. ~Kermaniz~.:~eh et al. technique uses MT fixative, see Ref. I. bEosinophils, mucous mast cells, serous mast ceils, dMay-Grunwald Giemsa staining. HDL Particles as the Trypanosome-killing Factor in Human Serum: An Exclusive or Inconclusive Role? In a recent Comment in Parasltology Today ~, Hajduk and colleagues describe the :~ndings by Tomlinson and co woH<ers 2 of 'HDL-independent killing of Trypc~nosorncJ brucei brucel by huma: serum' as new and surprising. Howecer, apart from the overdue recognition of trypanolytic activity in Tangier sera. which confirms our own experience (see below), we feel that Tomlinson's paper provides neither wholly new data on the molecular nature of the trypanolyti¢ factor, nor conclusive evidence for the existence of an entirely HDL-independent, second category of lyric particle. First, following the key observations of Yamaguchi and Rift<inthat trypanob'tic activity is associated with bulk seru~ n ~-lipoproteinslHDL, it was Barth 3 who originally isolated a cytotoxic ~ 10O0kDa particle from pooled human sera whose peptide composition was very different from HDL. In addition to unknown minor compon,~nts, we subsequently identified the main constituents of this particle as the p.-heavy chain of IgM (85 kDa) and K- and ~ -light chains of Ig (29 and 31 kOa, respectiwfiy). Although this work avoided 2SO fractionation by ultracentdfugation, and hence the generation of artefactual particles 4. it was not pursued furthe~ changing serum pools gave inconsistent findings, while Rifkin had already provided compelling evidence against IgM as 'the' trypanolytic factor. However, we later used gel filtration of individual sera employing fast-protein liquid chromatography (FPLC) to demonstrate that only certain individuals had particles with powerful activity in the ~ 1000 kDa size range of Barth's tr'/panocidal factors, ie. much larger than bulk serum HDL (I 80-400kDa): thls observation is now confirmed by Tomlinson's report 2, Second, the choice of HDL-deficient sera from Tangier patients would seem inappropriate, particularly as two key experiments were not reported. However, before debating this point, it appears essential to define HDL. This is not easy, given the marked heterogeneity of this class o f lipoproteins. Thus, as described elsewhere 4, although most HDL particles ah'e of 8-12 nm diameter, contain apoA-I (with or without apoA-ll) and float within a density range of 1.063-1.21 gml -i, minor subpopulations which do not fulfill all these criteria are welt recognized. Nevertheless, for the present purpose, we shall consider HDL as an apoA-l-containing particle with a density > 1.063 g ml- L notwithstanding the presence of"/-LpE 6 and apoA-ll-only 7 particles in both normal and dyslipoproteinaemic sera. The molecular defect of Tangier disease is not known but, unlike certain patients with familial apoA-I deficiency caused by mutations in the apoA-I gene, Tangier sera is not completely devoid of apoA-l: typically 2-3% is present (2-4mgdl -l) and HDL particles of diameter 5.5-7.5 nm, 20-25 nm and > 100 nm have been identified 8. Thus, in our study of normal sera, we still considered the ~ 1000 kDa trypanolytic fraction to be an HDL particle; it contained apoA-I along with other unidentified proteins and, importantly, the corresponding sera totally lost all lyric activity when treated with anti ~-lipoprotein antiserum s. By contrast, Tomlinson and colleagues 2 did not report this standard immuno- precipitation experiment 9 nor did they provide a distribution profile of apoA-I using sensitive, but routine, enzyme-linked immunosorbent assay (ELISA) or radioimmunoassay (RIA) detection. Rather, their argument for an HDL-independent lytic activity relied on the very low cholesterol content of their ~ 1000 kDa fraction and the apparent Parasitology Toda!4 voL 12, no. 6, •996