Transmission electron microscopy of ultrathin sections reveals multitubular pod inclusions in non-pathological cockroach granular hemocytes associated with skeletal muscle cells and tracheocytes. Each pod contains up to 39 simple (i.e., single-walled) or compound coaxial (i.e., doubleor triple-walled) tubules. Tubules are about 70-85 nm in diameter, depending on tubule complexity. The distribution of tubule types and tubule dimensions suggest that the outer tubule wall forms first, in the case of doubleor triplewalled tubules. Tubule lumens usually are empty. Insect hemocytes have been studied extensively (for recent reviews, see Gupta, 1979a, 1985). During a study of the ultrastructure of leg muscle in Periplaneta americana, we saw multitubular pod inclusions in a granular hemocyte closely associated with skeletal muscle cells and tracheocytes. A literature search revealed only three reports of the ultrastructure of these inclusions (Baerwald, 1979; Baerwald & Boush, 1970; Ennesser & Nappi, 1984). Because Baerwald (1979) and Baerwald & Boush (1970) described hemocytes collected by dripping hemolymph from a cut antenna end into fixative and because Ennesser & Nappi (1984) studied similarly collected hemocytes and hemocytes taken from an encapsulation response, we considered description of inclusions from an in situ and non-pathological granular hemocyte to be pertinent. MATERIALS AND METHODS A female of Periplaneta americana, about 6 cm in length, was captured in our greenhouse, chilled in a refrigerator for 15 min, and decapitated. The skeletal muscle from the femur was removed from the exoskeleton and immersed in a solution containing 2.5% glutaraldehyde buffered at a pH of 7.4 with 0.1 M phosphate buffer (Millonig, 1961). Primary fixation was carried out overnight (15 h) at 5?C. Four consecutive 15-min primary washes with buffer were followed by 2 h postfixation in 2% OsO4 in the same buffer. Four 15-min washes in the buffer followed. Dehydration was carried out by passage, at 10min intervals, through an ethanol series (1:1 50% ethanol: buffer, 50%, 75%, 85%, 95%, 100%, 100%, 100% ethanol), followed by three 10-min changes in propylene oxide. Overnight (12 h) infiltration was accomplished in a 1:1 mixture of propylene oxide and Epon LX-112 (Luft, 1961). A 1:3 mixture of propylene oxide and Epon was allowed to infiltrate the specimens for 8 h. Specimens were embedded in Epon in aluminum weighing dishes and subsequently placed in a polymerizing oven at 60?C for 48 h. Ultrathin (40-60 nm) sections were cut using a diamond knife (E.I. Dupont, Wilmington, Delaware) mounted on a Sorvall Porter-Blum MT2-B ultra-miTRANS. AM. MICROSC. Soc., 109(2): 168-173. 1990. ? Copyright, 1990, by the American Microscopical Society, Inc. This content downloaded from 207.46.13.162 on Thu, 30 Jun 2016 05:47:28 UTC All use subject to http://about.jstor.org/terms VOL. 109, NO. 2, APRIL 1990 crotome. Sections were stained with saturated uranyl acetate in 50% ethanol (Gibbons & Grimstone, 1960) followed with a 0.25% lead citrate solution (Venable & Coggeshall, 1965). Stained sections were examined with an RCA EMU2D transmission electron microscope operated at 50 kV. RESULTS AND DISCUSSION The terms "substructured inclusion" (Baerwald & Boush, 1970), "structured inclusion" (Baerwald, 1979; Ennesser & Nappi, 1984), "cylinder inclusion" (Baerwald & Boush, 1970; Ennesser & Nappi, 1984), and "structured granule" (Gupta, 1979b) have been applied to the objects of this report. We believe that "multitubular pod inclusions" is a more appropriate term inasmuch as "structured inclusions," "substructured inclusions," and "structured granules" include several different kinds of inclusions; e.g., those with "regular packed subunits" (fig. 7 of Baerwald & Boush, 1970) and those with "a series of band-like subunits" (fig. 5 of Baerwald & Boush, 1970). Furthermore, "cylinder" denotes either a solid or a hollow structure. These inclusions do resemble a pod composed of many tubular (i.e., hollow cylindrical) substructures. Incidentally, although considerable variation exists in hemocyte classification, we accept the term "granular hemocyte," used by Arnold (1972) in his comparative study of the hemocytes of cockroaches, for the host cell of the inclusions here described. The multitubular pod inclusions occurred in about 5% of the granular hemocytes studied by Baerwald & Boush (1970) and only in the experimental animals studied by Ennesser & Nappi (1984), who also found them extracellularly among cellular debris. We found them in three of about 50 different granular hemocytes examined. The earlier reports state that each inclusion is bounded by a unit membrane (Baerwald, 1979) surrounding 9-80 hollow cylinders, each with a diameter of about 800 A, a length of about 0.8-1.5 ,A, and a cylinder wall of the size of a unit membrane, but of a different structure (Baerwald & Boush, 1970). Baerwald & Boush (1970) reported single-, double-, and triple-walled cylinders, with the double-walled most numerous. Ennesser & Nappi (1984) reported only singleand double-walled cylinders. We believe the terminology requires some clarification. The tubules within a single pod are simple (single-walled) or a mixture of simple and compound (doubleor triple-walled). The compound tubules may be considered to be composed of two or three coaxial tubules, each with its own wall. This latter interpretation is necessary if one is to accept the statement of Baerwald & Boush (1970) that the cylinder wall is of the size of a unit membrane; i.e., each of the coaxial tubule walls has the size of a unit membrane. We refer, then, to tubules as single-, double-, or triple-walled. Our observations of the inclusions used to illustrate this report are derived from numerous serial and adjacent sections through a single elongated granular hemocyte which was closely applied to the surface of several skeletal muscle cells and tracheocytes (Fig. 1). Subsequent sections of about 50 different hemocytes revealed only two with pod inclusions. In each of these two hemocytes, 169 This content downloaded from 207.46.13.162 on Thu, 30 Jun 2016 05:47:28 UTC All use subject to http://about.jstor.org/terms TRANS. AM. MICROSC. SOC.