Naturwissenschaften 77,189- 190 (1990) © Springer-Verlag 1990 Phylogeny of Arthropod Immunity An Inducible Humoral Response in the Kalahari Millipede, Triaenostreptus triodns (Attems) E. van der Walt and E. McClain Department of Physiology, Medical School A. Puren and N. Savage Department of Medical Biochemistry, University of the Witwatersrand Medical School, Parktown 2193, Johannesburg, South Africa Humoral immunity has been dem- onstrated in a number of different in- vertebrates, for example earthworms, crabs and snails [1]. However, amongst the arthropods work has focussed on more highly evolved short-lived species of insects such as the silk worm, wax moth, fruit fly and others [2, 3]. Un- fortunately, information on how ar- thropods other than insects respond to bacterial infection is limited. Long- lived terrestrial arthropods like scorpions and millipedes may have evolved different mechanisms for cop- ing with pathogens. This paper reports on an inducible anti- bacterial response to gram-negative bacteria (E. colt) in the Kalahari mil- lipede, Triaenostreptus triodus (Diplo- poda: Spirostreptidae). The potent humoral response to infection appears to be similar to that found in short- lived insects. Although T. triodus produces several proteins in response to bacterial challenge as illustrated by po- lyacrylamide gel electrophoresis, only one protein expressed bactericidal ac- tivity when subjected to the gel overlay. This is the first report of a humoral re- sponse in a long-lived diplopod. In- ducible proteins responsible for anti- bacterial activity in arthropods may therefore have a long phylogeny. The Kalahari millipede (Fig. 1) is one of the most successful arthropod species in the arid regions of South Africa, Botswana and Namibia. They are espe- cially abundant in the Kalahari desert after rain. Here, these long-lived "Son- gololos" feed on the newly sprouted vegetation, especially the leaves and yellow flowers of Tribulus sp., so com- monly covering the Kalahari sands. They also feed on detritus, algae, fungi and a myriad of small mammal scat Naturwissenschaften 77 (1990) found on the desert floor. After surface activity of about 4 months they return underground to await next year's rains. Certainly their life style and feeding habits are conducive to exposure and attack from bacteria. In a North American millipede, Orthoporus or- natus microbial mortality has been ob- served [4]. T. triodus is an ideal animal in which to study immune responses in the laboratory, as aside from its life style possibly replete with pathogens, large quantities of hemolymph can be sampled from the same individual with- out harm. Millipedes were collected at Nossob in the Kalahari Gemsbok Park, South Af- rica (20°30"N; 21°50"E). The mil- lipedes were kept at room temperature (24°C) in glass aquaria on Kalahari sand and fed on a diet of rolled oats and lettuce. Because of the reported synthesis of wound-healing proteins upon injection of any pathogen [5] and the necessity of using pyrogen-free water and saline to make up all solu- tions, non-injected as well as saline-in- jected millipedes were used as controls. Fig. 1. The Kalahari millipede Triaenos- treptus triodus, showing both sampling and injection site (arrowed) © Springer-Verlag 1990 Saline-injected controls often produced proteins that were slightly positive on the gel-overlay method, a phenomenon also noted by other workers [6]. The bacterium used was the gram-negative Escherichia eoli K12. This organism is commonly used in immune studies. The bacteria were cultured in 93 x 15 mm sterile Petri dishes containing either Nutrient Agar or McConkey Agar (Merck, Germany). Bacterial suspen- sions for injection were made up by scraping the bacteria off the culture plates with a heat-sterilized spatula and suspending them in insect saline (0.7 % w/v NaC1). The bacterial suspensions were counted under a microscope in an Improved Neubauer Counting Chamber (Spencer, Brightline, USA) under oil. Hemolymph was collected by inserting a 26G hypodermic needle be- tween two adjacent segments in the mid-dorsal region (Fig. 1). Hemolymph was drained into an Eppendorf tube kept on ice. Samples were taken at va- rious time intervals after injection (24, 48, 72, 96 h). This was done to de- termine the maximum temporal syn- thesis of antibacterial proteins. To de- termine whether any antibacterial activ- ity was induced by the injection of E. coli the plate assay technique was first employed [7]. Specific antibac- terial activity in the proteins separated by acidic electrophoresis was detected by means of a gel overlay [3, 8]. Vac- cinated and control hemolymph samples were separated towards the cathode in a 10% non-denaturing acidic polyacrylamide gel (pH 4), in duplicate. After electrophoresis one gel was incubated in nutrient broth with 0.2 M potassium phosphate buffer pH 7.4 for 1 h. The gel was then overlayed with nutrient agar which was seeded with 105 bacteria E. coli K12 per ml. The gel overlay was incubated at 35 °C for 24 h to develop zones of inhibition [3]. The duplicate gel run under identical conditions was stained to detect the proteins. When subjected initially to the plate as- say method, growth of E. coli was un- affected by control hemolymph (Fig. 2A). However, hemolymph taken from a millipede 48 h after infection inhib- ited E. coli growth (Fig. 2B). This is seen as a clear zone of inhibition whose magnitude appears similar to that found for various vaccinated insect species subjected to this method [6, 7]. 189