Maternal and Paternal Nitrogen Investment in Blattella Germanica (L.) (Dictyoptera; Blattellidae)

Insects utilize a variety of strategies and adaptations to fulfil nutritional requirements for their growth, development and reproduction (Thornhill and Gwynne, 1986; Slansky and Rodriguez, 1987). Reproductive strategies among cockroaches are diverse and may include oviparity, ovoviviparity or viviparity (Roth, 1968; Stay and Coop, 1974), prezygotic investment at mating by the male (Mullins and Keil, 1980; Schal and Bell, 1982) and maternal or biparental care (Nalepa, 1988). Cockroaches also possess a considerable amount of latitude in their food selection (Lembke and Cochran, 1989), partly because of their ability to mobilize stored fat-body urates. The mobilization of these internally stored fat-body urates is probably achieved by bacterial symbionts, which provide a resource of nitrogen-containing materials for these insects (Mullins and Cochran, 1987; Wren and Cochran, 1987; De Piceis Polver et al. 1986) and allow them to survive on a variety of food materials, including those that may be low in nitrogen (Cochran, 1985). The dietary requirements of reproducing cockroaches indicate that dietary nitrogen may influence fecundity (Melampy and Maynard, 1937; Haydak, 1953; Gordon, 1959; Kunkel, 1966; Mullins and Cochran, 1975a,b; Cochran, 1983; Durbin and Cochran, 1985; Hamilton and Schal, 1988; Hamilton et al. 1990). It has also been reported that uric acid may be transferred from males to females at mating, presumably to support the nutritional requirements of oogenesis and/or embryogenesis (Mullins and Keil, 1980; Schal and Bell, 1982). To evaluate dietary influences on parental nitrogen investment into German cockroach progeny, two radiolabelled materials; purines (as [¹⁴C]hypoxanthine) and an amino acid (as [³H]leucine) were injected simultaneously into either females or males maintained on two different dietary nitrogen levels. Dual radiolabel techniques were used to determine incorporation, retention and transfer of radiolabel between males and females and their progeny. In a separate experiment, a stable nitrogen isotope (¹⁵N) of uric acid was provided in a diet to determine the fate of urate nitrogen in ootheca-producing females.

German cockroaches used in this study were obtained from established cultures held at approximately 25°C and ambient humidity in 41 glass battery jars and provided with dog food (25 % crude protein) and water ad libitum. The sexes were separated at a late nymphal stage and held in groups until eclosion in 11 glass jars also provided with food and water. At adult eclosion, single females were placed in 10 cm x 2 cm plastic Petri dishes fastened to the top of a 1.8 g vial containing distilled water. A glass tube (0.3 cmx7 cm) packed with a cotton wick was inserted through the top of the vial and the bottom of the Petri dish, providing access to water. The diets used were: a 5 % protein diet described previously (Mullins, 1974), a 5 % protein diet containing 10 % [¹⁵N]uric acid, or dog food (25 % crude protein). They were dispensed into small containers, packed, steamed (10min) and oven-dried (80°C for 1 h) prior to presentation to the insects. Older nymphs and adults were held in an environmental chamber (28±2°C, 55-65% relative humidity). Data were analyzed using the Student’s t-test (Sokal and Rohlf, 1981).

A motor-driven microapplicator was used to inject 1 μl volumes of radiolabelled [³H]leucine (L-[4,5-³H]leucine, specific activity 52 Ci mmol^-1; ICN-Pharmaceuticals) and [¹⁴C]hypoxanthine ([8-¹⁴C]hypoxanthine, specific activity 53.4mCi mmol^-1; ICN-Pharmaceuticals) into the insect’s abdomen. Radiolabelled hypoxanthine was used as the urate (purine) source instead of xanthine or uric acid because its greater solubility in aqueous solutions facilitated the injection of relatively equivalent amounts into each insect. The subsequent in vivo conversion of hypoxanthine to uric acid is rapid and quantitative (Mullins and Keil, 1980; Cochran and Mullins, 1982; Engebretson and Mullins, 1983, 1986).

At the conclusion of radiolabel experiments, insects were frozen and whole bodies and oothecae were homogenized in 1 ml of 0.6 % Li₂CO₃ (80°C for 10 min) and centrifuged at 1000g (10min). Samples of extracts (250-500 μl) were placed into 8 ml phase-combining system (PCS) (Amersham Searle)/xylene (1:1) scintillation cocktail. Separation and quantification of radiolabelled urate were by thin-layer chromatography (TLC) on 1mm cellulose plates (20 cm X 20 cm) developed in n-butanol/methanol/water/NH₄OH (60:20:20:1) and scintillation counting of the recovered and extracted urate spots, respectively. Total urate was also determined by the uricase method (Dubbs et al. 1956) as modified by Mullins and Cochran (1976). Additional details of the three radiolabel experiments are provided in table footnotes.

Newly eclosed adult females were provided with a 5 % protein diet containing 10% [¹⁵N]uric acid ([l,3-¹⁵N]uric acid, 95 atom%; Merck & Co.). When 19mg of this diet had been consumed, insects were placed on either a dog food or a 5 % protein diet, and were provided with an opportunity to mate (placed with two males for 1 h daily). Oothecae were removed and frozen 21 days after they appeared. Because of the expense of analysis, only one female and her ootheca from each diet were analyzed. The presence of ¹⁵N in the amino acids in the oothecae was determined by hydrolysis with 6 mol 1^-1 HC1 under N₂ at 110°C for 24 h. The hydrolysate was evaporated to dryness with a stream of N₂, dissolved in water, filtered and again evaporated to dryness under a stream of nitrogen. The resulting crude amino acids were converted to n-butyl esters by heating for 3 h at 100°C with 3 mol 1^-1 HC1 in n-butanol. After removing the HC1 and n-butanol by evaporation under a stream of nitrogen, the resulting amino acids were converted to their trifluoroacetic acid derivatives by reaction for 12 h with a mixture of trifluoroacetyl anhydride and methylene chloride (1:1 v/v). Evaporation of the solvent left a brown residue which was mixed with 3 mol 1^-1 NaHCO₃, extracted with heptane and dried over Na₂SO₄. The presence of ¹⁵N in the individual amino acid derivatives contained in the concentrated heptane solution was determined by gas chromatography-mass spectroscopic analysis, as previously described (White, 1985).

Determinations of whole-body and oothecal nitrogen and urates in non-radiolabelled insects were carried out on lyophilized tissue pulverized using a dental amalgamator and liquid nitrogen. Total nitrogen was determined using a micro-Kjeldahl procedure and urate content was estimated using a spectrophotometric uricase assay (Mullins and Cochran, 1976).

The total mass, the nitrogen content and the urate nitrogen content of reproducing females were obtained to provide information on female allocation of nutritional resources into progeny. Dog-food-fed females plus their newly eclosed nymphs and emptied oothecae had a mean mass of 36.5±2.1mg and contained a total of 235±49μmol nitrogen (Table 1). Urate nitrogen accounted for 51% (females) and 6.4% (nymphs) of the total nitrogen (Table 1).

Radiolabelled leucine was injected into females 0–2 or 5–7 days after imaginal eclosion to determine the effects of diet and time of injection on leucine-derived radiolabel incorporation into newly formed oothecae. Females on both diets retained more radiolabel when they were injected 0–2 days after imaginal eclosion (Table 2). Radiolabel incorporation into oothecae by females maintained on the dog food diet and injected 5–7 days post-imaginal eclosion was greater than that of females injected at 0–2 days, but the difference for females fed the 5 % protein diet was not significant. Similarly, differences were observed in total radiolabel retention in dog-food-fed females and their oothecae injected at either 0–2 or 5–7 days, but not in the females fed a 5 % protein diet.

Several observations can be made regarding diet effects on [³H]leucine retention/incorporation. Females injected 0-2 days after imaginal eclosion and maintained on either diet retained similar amounts of radiolabel. However, more radiolabel was incorporated into oothecae by females maintained on the low-protein (5 %) diet. Females on both diets injected 5-7 days after imaginal eclosion also retained similar levels of radiolabel, but more radiolabel was incorporated into oothecae by females maintained on the higher-protein (dog food) diet.

A dual-label experiment was conducted using two metabolically different nitrogen-containing resources to examine the effects of diet on their incorporation during oocyte development to maturation. [¹⁴C]Hypoxanthine was used to examine incorporation, translocation and metabolism of urate and [³H]leucine was used to assess the incorporation and metabolism of a representative amino acid (Table 3). The time required to produce an ootheca by females maintained on the 5 % protein diet was significantly longer than that required for females on the dog food diet (42.2 and 13.9 days, respectively), indicating that lower dietary nitrogen levels, beginning 14 days prior to eclosion but not 7 days prior to eclosion, affected the rate of oothecal production (Table 2, footnote).

Females maintained on dog food retained 25% of the injected [³H]leucine-derived radiolabel in their bodies and invested 31 % into their oothecae (Table 3). Total retention of the radiolabel to the time of oothecal production was 56%. Females maintained on the 5 % protein diet retained a similar amount of radiolabel in their bodies but significantly less radiolabel was incorporated into their oothecae (17 %). This lower rate of leucine incorporation suggests that more of the amino acid was metabolized by females as a result of their maintenance on the lower-protein diet.

The fate of radiolabelled hypoxanthine also indicates that differences in allocation occurred in response to dietary regimen. Females on dog food retained 65 % of the injected dose as [¹⁴C]hypoxanthine-derived radiolabel and invested 29% in their oothecae (Table 3). Females maintained on 5 % protein retained significantly less radiolabel (12 %) but invested a similar amount of radiolabel into their oothecae (41%). Total hypoxanthine-derived radiolabel retention was higher in the dog-food-fed females and their oothecae than in those maintained on the lower protein diet.

Additional information on the fate of urates in the dual-label experiment is provided in Table 4. In the females fed dog food, 98% of the hypoxanthine-derived radiolabel remaining in their bodies was [¹⁴C]urate, as was 84% of the radiolabel incorporated into their oothecae. Those females maintained on 5 % protein also retained 98 % of the hypoxanthine-derived radiolabel in their bodies as [¹⁴C]urate, but only 75 % of the radiolabel incorporated into their oothecae was [¹⁴C]urate. Females on the dog food diet contained over four times as much urate as did those on the 5 % protein diet. Oothecae from females fed dog food also contained more urate than those on the 5 % protein diet. Significant quantities of urates were incorporated into oothecae during oogenesis/oothecal production.

[¹⁵N]Uric acid was fed to females on dog food and 5 % protein diets to obtain preliminary information on the fate of uric acid nitrogen during oothecal production and embryonic development. Both 21-day-old oothecae contained the following ¹⁵N-labelled amino acids; alanine, proline, glutamic acid and aspartic acid. Comparison of the relative amounts of these amino acids indicates that the ootheca obtained from the female maintained on the 5 % protein diet contained more of the ¹⁵N label (Table 5).

A dual radiolabel protocol using [¹⁴C]hypoxanthine and [³H]leucine was employed to assess the incorporation of radiolabel injected into males and subsequently transferred to females and their oothecae after mating. As in the maternal investment dual-label experiment, [¹⁴C]hypoxanthine was used as an indicator of the transfer of urate-derived metabolites and [³H]leucine was used as an indicator of the transfer of leucine (amino acid)-derived metabolites to females and their oothecae. Males injected with the two radiolabelled compounds were paired with females and the distribution of radiolabel was determined after an ootheca had been produced. The times required to produce one ootheca by females maintained on dog food prior to pairing, but placed on either dog food or 5% protein diets at pairing, were similiar (Tableó, footnote). Male radiolabel retention is indicated in two categories (mated and unmated). Those males from each replicate group (2 males and 1 female) that had the lower ¹⁴C content of the pair were designated as having mated, since radiolabel contained in the accessory gland as [¹⁴C]urate is voided at mating with the spermatophore. Correspondingly, those males that had the higher ¹⁴C content were considered to have remained unmated.

Unmated males maintained on either diet retained significantly more ¹⁴C radiolabel than mated males (Table 6). No differences in ¹⁴C content between the unmated males on either of the diets were observed but dog-food-fed mated males retained more ¹⁴C radiolabel than 5 %-protein-fed mated males. Females on both diets retained less than 2 % of the ¹⁴C radiolabel injected into the males. However, the oothecae of the 5 %-protein-fed females contained over 10 times more than the dog-food-fed females of the ¹⁴C radiolabel provided by the males. Total female and oothecal ¹⁴C radiolabel acquisition values indicate that less than 2 % of the radiolabel injected into males was contained in dog-food-fed females and their oothecae whereas over 8 % of the radiolabel injected into males was contained in 5 %-protein-fed females and their oothecae.

Unmated and mated males maintained on either of the two diets retained similar amounts of ³H radiolabel ranging from 31 to 35 % of the injected dose (Table 6). Differences were found in ³H content of females maintained on the two diets (dog food diet 0.5 %, 5 % protein diet 1.8 % of the radiolabel injected into the males). Similar differences were observed in the ³H content of oothecae obtained from these females. Females maintained on the 5 % protein diet contained more than three times as much ³H in their bodies and oothecae as those maintained on the dog food diet.

Fig. 1 provides comparisons of the total ¹⁴C- and ³H-derived radiolabel retention in the males and acquisition by the females maintained on the two diets. Differences in total ¹⁴C and ³H radiolabel retention between unmated and mated males were used to estimate the amount of total radiolabel (¹⁴C+³H) that was made available to females by males during mating. Percentages in this figure represent the relative amount of material that females acquired and retained from males during mating. Comparison of total female ¹⁴C and ³H content in disints min^-1 demonstrates that those females maintained on the 5 % protein diet retained more of the radiolabel than those females maintained on dog food. The amount of oothecal radiolabel content obtained from the male shows a striking difference between diets (5 % protein diet 39 % ; and dog food diet 8 %). All of the radiolabel found in mated females and their oothecae was contributed by the mated males (5 % protein diet 63 %, dog food 17 %).

Parental investment generally does not scale linearly with species mass because larger species invest less material into their progeny relative to their body mass per unit time (Reiss, 1985). Kunkel (1966) has compared reproductive potential in two cockroach species, reporting that female Periplaneta americana (larger species; live mass approximately 1200 mg) diminished their food reserves by 15 % with each ootheca produced, but female B. germanica (live mass approximately 105 mg) used up almost all (90%) of their food reserves accumulated during the preovipositional period on a single ootheca. We found that significant amounts of the dry mass (34 %) and of the nitrogen content (26 %) of ootheca-carrying female B. germanica were contained in their newly eclosed oothecae (Table 1).

Insect nutritional status is known to affect reproductive performance (McCaffery, 1975; Slansky, 1980; Slansky and Rodriquez, 1987). Food deprivation in B. germanica causes a delay in the reproductive cycle (Kunkel, 1966; Mueller, 1978; Durbin and Cochran, 1985). During the preovipositional period, Cochran (1983) found that feeding and drinking in B. germanica females was at a high level. This fell abruptly at oviposition, with females feeding and drinking only sparingly while carrying an ootheca. Food quality has been shown to influence diet consumption. Female B. germanica compensate for low dietary nitrogen content (5 % protein diet) by consuming more of the diet (Hamilton and Schal, 1988).

We found that the time between post-imaginal eclosion and oothecal formation increased when late instar nymphs were placed on low-protein diets for over 7 days. Late instar nymphs placed on the 5 % protein diet 7 days before imaginal eclosion showed no dietary differences in the time required for oothecal production. However, females placed on the 5 % protein diet 14 days before imaginal eclosion required more time for oothecal production. Hamilton and Schal (1988) found no differences in the time required for oothecal formation between females placed on a 5 % protein diet and those fed on rat food (23 % protein) at imaginal eclosion. However, another cockroach species (Supella longipalpa) maintained on low-protein diets from the nymph to adult stage required more time to produce an ootheca (Hamilton et al. 1990). Prolonged maintenance of pre-imaginal nymphs on low-protein diets appears to delay oothecal formation.

Vitellins and urates are two nitrogen-containing materials available during development of B. germanica embryos. Females provision their eggs with vitellins during oocyte development, providing a major amino acid resource for embryogenesis. Their utilization is correlated with embryonic growth (Purcell et al. 1988). Leucine is a major vitellin constituent, representing 8.7mol% of the amino acid content in the vitellogenin precursor (Kunkel and Pan, 1976). Information obtained from [³H]leucine injection into females (Tables 2 and 3) indicates that females maintained on different dietary nitrogen levels retained similar amounts of [³H]leucine-derived radiolabel. However, females on the low-protein diet (5 % protein) produced oothecae containing less [³H]leucine-derived radiolabel. This reduction may reflect a general decline in radiolabel available for oogenesis in the amino acid pools of females placed on the 5 % protein diet 14 days before imaginal eclosion.

Fat-body urates appear to play an important role in cockroach nitrogen metabolism (Cochran, 1985; Mullins and Cochran, 1987; Wren and Cochran, 1987). Whole-body urate levels can be correlated with dietary nitrogen status and [¹⁴C]urates are metabolized by cockroaches maintained on variable-nitrogen diets (Mullins and Cochran, 1987; Engebretson and Mullins, 1986). It also appears that oothecal urates are utilized during embryogenesis because their levels decline as the embryos develop (Mullins and Keil, 1980; Cochran and Mullins, 1982).

In this study, urate retention in females and its transfer to their oothecae was influenced by diet. It appears that females on the lower-nitrogen diet transfer more of their available urates into their oothecae. Females fed dog food retained more unlabelled and radiolabelled urate in their bodies than those on the 5 % protein diet. Oothecae from dog-food-fed females contained significantly more radiolabel (primarily as urates) than the female, but females on the 5 % protein diet contained significantly less ¹⁴C radiolabel than their oothecae (Table 3).

Specific information on cockroach urate catabolism is not available (Wren and Cochran, 1987). However, significant quantities of ¹⁴CO₂ may be released in cockroaches provided with [¹⁴C]urate stores (Cochran and Mullins, 1982; Engebretson and Mullins, 1986). The lower levels of [¹⁴C]urate and total urate in the females fed on the 5 % protein diet can be attributed to increased metabolism of the stored urates, which is known to occur under circumstances of low dietary nitrogen (Table 4; Mullins and Cochran, 1987). Eighty-four per cent of the radiolabel contained in the oothecae obtained from dog-food-fed females and 75 % of the radiolabel contained in oothecae obtained from females fed on the 5 % protein diet was attributable to [¹⁴C]urate. Because ¹⁴C was largely retained as urates in the females, 98 % in all females, whether fed on dog food on the 5 % protein diet, urates transferred to oothecae are metabolized to a greater extent during embryogenesis.

Radiotracer studies using [¹⁴C]purine intermediates are useful only in studies of pathways involving carbon atoms but do not necessarily reflect the path of nitrogen contained in the urates. We conducted a preliminary experiment using a stable isotope of nitrogen ([¹⁵N]urate) to determine whether ¹⁵N could be incorporated into nitrogenous metabolites. We indentified four amino acids obtained from 21-day-old oothecae that contained measurable levels of ¹⁵N enrichment. The ootheca obtained from a female maintained on the 5 % protein diet contained larger amounts of these ¹⁵N-enriched amino acids than the ootheca from the female fed on dog food (Table 5). This information supports the hypothesis that urate-derived nitrogen is assimilated into nitrogen-containing compounds and that this mobilization is more pronounced in cockroaches on low-protein diets.

Female metabolism and utilization of urate metabolites or direct urate incorporation into oothecae is indicated by several lines of evidence. These include: (1) steadily decreasing urate levels during embryogenesis (Mullins and Keil, 1980) and fluctuations in biréfringent ‘white’ granules, assumed to be urates, during various stages of embryogenesis (Tanaka, 1976); (2) the presence of ¹⁴C radioactivity in tissue extracts that was not attributable to [¹⁴C]urate (Mullins and Keil, 1980; Cochran and Mullins, 1982; this study), and (3) the finding that ¹⁵N derived from [¹⁵N]urate can be found in several amino acids (this study).

Much evidence has accumulated over the years implicating cockroach fat-body endosymbionts (bacteriocytes) in the metabolism of stored urates (Cochran, 1985; Wren and Cochran, 1987). These bacteriods are transmitted to oocytes during oogenesis (Sacchi et al. 1985) and are recognizable 10 days after the onset of embryogenesis (Tanaka, 1976). It is likely that they are involved in urate metabolism in the embryos. Nutritional symbiosis involving urate metabolism has been demonstrated in termites (Potrikus and Breznak, 1980a,b) and in molgulid tunicates (Saffo, 1988). In termites, uricolysis is anaerobic (Breznak, 1982). Wren and Cochran (1987) suggested that anaerobic uricolysis occurs in cockroach bacteriocytes.

Prezygotic paternal investment may include indirect nutritional contributions to progeny provided to females by males during mating (Zeh and Smith, 1985). Nutrient contributions may be derived from male accessory glands and/or emptied spermatophores consumed by females (Leopold, 1976; Thornhill, 1976; Markow and Ankney, 1984; Sakaluk, 1984). Zeh and Smith (1985) noted that physically harsh or biotically dangerous habitats and ephemeral, highly prized, productive resources may be associated with high levels of paternal investment. Some information is available among insects to support this hypothesis. In Caryedon serratas (Coleoptera; Bruchidae), male secretions may constitute an important trophic contribution that increases female fecundity and vitellogenesis (Boucher and Huignard, 1987). Interspecific mating differences between Drosophila melanogaster and D. mojavensis suggest that because D. mojavensis lives in a harsh environment, often with limited resources, it remates more frequently in order to obtain nutrients from male ejaculates (Markow and Ankney, 1984). Cochran (1985) has suggested that, because of the geological period in which cockroaches evolved, the availability of dietary nitrogen was scarce or unpredictable. As a result, cockroaches may have been nitrogen scavengers. A portion of this strategy has been reviewed with respect to female urate (nitrogen) metabolism in a preceding section of this discussion. It appears that the males of many cockroach species may also be involved in nitrogen scavenging through uric acid found in association with the male accessory glands of some cockroach species (Roth and Dateo, 1964; Roth, 1967). Radiolabelled urates from male accessory gland urates are transferred to females in B. germanica (Mullins and Keil, 1980) and in Xestoblatta hamata (Schal and Bell, 1982). Schal and Bell (1982) found that female X. hamata feed on urates offered by males after copulation and that females on nitrogen-deficient diets ingest and transfer more male-derived urates to their maturing oocytes than do females on high-protein diets. Blattella germanica males provide more [¹⁴C]urates to females on 5 % protein diets than to females on dog food, a major portion of which is then transferred to their oothecae (Mullins and Keil, 1980; Table 2, this study). Similarly, more [³H]leucine-derived materials are transferred from the males to females with subsequent incorporation into their oothecae. The source of these [³H]leucine-derived materials has not yet been established, but they are probably derived from spermatophores or seminal fluids. The nature of the materials contributed by males appears to be related to dietary nitrogen levels. Mated males fed on dog food provided about 13% of their [³H]leucine-derived radiolabel and 20 % of the [¹⁴C]hypoxanthine-derived radio-label (as urates) to females. In contrast, males fed on 5 % protein contributed 1 % of their [³H]leucine-derived radiolabel and 41 % of their [¹⁴C]hypoxanthine-derived radiolabel. Males on the low-protein diet apparently utilized the ³H-labelled amino acid for their own short-term metabolic needs while providing more [¹⁴C]urate to females. Females on the 5 % protein diet transferred ten times as much [¹⁴C]urate into their oothecae as did those on dog food diets.

Several observations can be made in interpreting the results of the experiments in this study. First, the sizes of the metabolic pools in both male and female B.germanica are undoubtedly influenced by the diet to which they have access. Introduction of radiolabelled materials into metabolic pools of differing sizes will affect the specific activity of those materials in these pools. Thus, when the radiolabelled materials are utilized or transferred, the results obtained will reflect both the specific activity of these materials and the net amount of the materials transferred. Second, information provided in Fig. 1 can be used as an indicator of the significance of two radiolabelled resources injected into males which were then contributed by the males to the females and subsequently transferred by these females into their oothecae. These two nitrogen-containing materials are representative of a variety of compounds that could be transferred in a similar manner. The contribution obtained by females that is subsequently invested into their oothecae can be used as an indicator of the significance of paternal investment. Finally, males can play a significant role in providing nutritional support to their progeny, particularly under circumstances where the female diet is nitrogen-deficient. This is supported by the finding that females on the 5 % protein diet incorporated more (63%) of the total radiolabel made available to them by males than did those on the dog food diet (17 %) (Fig. 1).

We thank Drs D. G. Cochran, C. A. Nalepa and M. H. Ross for critical review of this manuscript. This research was supported in part by the Virginia Agricultural Experiment Station, Hatch project no. 130929.