Cysteine eliminates the feeder cell requirement for cultivation of Trypanosoma brucei bloodstream forms in vitro.

In all previous studies, bloodstream forms of Trypanosoma brucei could be grown in vitro only when supported by a feeder layer of mammalian fibroblasts. We have axenically cultivated bloodstream T. brucei by adding L-cysteine at regular intervals and appropriate concentrations. The optimum cysteine concentration depends on cell density and is close to physiological serum levels. At concentrations greater than 24 mg/liter (2 X 10(-4) M), cysteine was acutely toxic to trypanosome concentrations of 3 X 10(7)/ml. Toxicity was prevented by addition of pyruvate or catalase, which neutralize H2O2 produced by cysteine autoxidation. In uptake studies using [35S]cysteine and [35S]cystine, T. brucei efficiently incorporated only cysteine. The Km for cysteine uptake was 4 X 10(-4) M. Cystine supported axenic growth if low concentrations of 2-mercaptoethanol were added at regular intervals.

Glycosyl-sn-1,2-dimyristylphosphatidylinositol is the membrane anchor for Trypanosoma equiperdum and T. (Nannomonas) congolense variant surface glycoproteins.

We have analysed the structures of the Trypanosoma (Nannomonas) congolense and T. equiperdum variant surface glycoprotein (VSG) membrane anchors. Myristic acid uptake, phospholipase treatment, and nitrous acid deamination showed that, for each species, the anchor is glycosyl-sn-1,2-dimyristylphosphatidylinositol, as has been previously described for T. brucei. Osmotic lysis of these trypanosomes resulted in the release of soluble VSG, lacking fatty acid. In both species and in T. evansi, an endogenous phospholipase C, which cleaved diacylglycerol from membrane form VSG, was identified.

Analysis of antigen switching rates in Trypanosoma brucei.

Previously quoted figures for the frequency of antigen switching in Trypanosoma brucei are based on incorrect assumptions. In order to determine the correct switching frequency, an equation was derived that takes the growth rates of the newly expressed antigen types into consideration as well as the proportion of switched trypanosomes and the number of generations since the population was antigenically homogeneous. When this equation was applied to published in vitro data, variable values were obtained for the switching frequency in clonal populations originally expressing one antigen type. The calculated most likely switching frequencies ranged from 1.4 X 10(-7) to 3.5 X 10(-6). This variation was probably caused by differences in the growth rates of the new antigen types in the population and failure to detect slow growing variants. To overcome these problems, an experimental procedure was developed to analyse the switching frequency in vitro. Trypanosomes were cloned and grown in parallel cultures. After an appropriate number of generations, cells expressing the original antigen type were destroyed and, from the proportion of cultures that contained new antigen types, the switching frequency was calculated. The technique minimized subculturing or other procedures that could distort the results. Although the method was optimized for analysing switching frequency, the values differed between experiments, ranging from 2.2 X 10(-7) to 2.6 X 10(-6) for one variant. Possible causes for the variations in switching frequency are discussed.

Biosynthesis of Trypanosoma brucei variant surface glycoproteins. N-glycosylation and addition of a phosphatidylinositol membrane anchor.

The variant surface glycoproteins (VSGs) of Trypanosoma brucei are synthesized with a hydrophobic COOH-terminal peptide that is cleaved and replaced by a glycophospholipid, which anchors VSG to the surface membrane. The kinetics of VSG processing were studied by metabolic labeling with [35S]methionine and [3H]myristic acid. The COOH-terminal oligosaccharide-containing structure remaining after phospholipase removal of dimyristyl glycerol from membrane-form VSG could be detected serologically within 1 min of polypeptide synthesis in two T. brucei variants studied. Addition of the oligosaccharide-containing structure was resistant to tunicamycin. VSGs synthesized in the presence of tunicamycin displayed lower apparent molecular weights, consistent with the complete inhibition of N-glycosylation at one (variant 117), two (variant 221), or at least three (variant 118) internal asparagine sites. In most experiments, N-glycosylation appeared to occur during or immediately after polypeptide synthesis but in a few cases N-glycosylation was delayed or incomplete. In all cases, addition of the COOH-terminal oligosaccharide-containing structure occurred normally. In dual-labeling studies, cycloheximide caused rapid inhibition of both [35S]methionine and [3H]myristic acid incorporation, suggesting that myristic acid addition also occurs immediately after polypeptide synthesis. Our data suggest that the complex ethanolamine-glycosyl-dimyristylphosphatidylinositol structure of membrane-form VSG is added en bloc within 1 min of completion of the polypeptide.