Inheritance of resistance to fenthion in Culex pipiens fatigans Wied*

Bull. Org. mond. Sante 11970, 43, 727-734 Bull. Wld Hlth Org. Inheritance of Resistance to Fenthion in Culex pipiens fatigans Wied. * CLAIRE DORVAL1 & A. W. A. BROWN2 The insecticide fenthion, which shows great promise for controlling the filariasis vector Culex pipiens fatigans, has been found in the laboratory to induce a certain amount of resistance. As the biochemical nature offenthion resistance was known to be due to an increase in a particular esterase activity, the inheritance of the character was investigated in the present study. It wasfound to be mainly due to a single slightly recessive gene asso- ciated with linkage-group 2, and the increase in detoxifying esterase activity showed a similar inheritance. The resistance to the insecticide fenthion-0,0- dimethyl O-[(4-methylthio)-m-tolyll phosphoro- thioate-induced in a Rangoon strain of Culex pipiens fatigans has been shown to be due to an in- crease in detoxification by esterase activity acting on fenthion and the fenoxon produced therefrom (Stone & Brown, 1969). The present investigation concerned the inheritance of the fenthion-resistance character, its linkage with marker genes, and the parallelism between the genetic factor(s) and the amount of esterase activity. MATERIAL AND METHODS The following strains were continuously reared in the laboratory on a larval diet of 5 parts brewers' yeast, 2 parts blood albumin and 1 part RNA, with adult blood meals on white mice: S: a susceptible Rangoon strain of Culex pipiens fatigans derived from a collection of females made in 1963 (LC50 o.0o24 ppm fenthion). R: a resistant strain developed from the above by 3 selections with DDT followed by 42 selections with fenthion (Stone & Brown, 1969), and maintained thereafter * From the Department of Zoology, University of Western Ontario, London, Canada. This investigation was supported by grants from the World Health Organization and the National Research Council of Canada. 1 Present address: Department of Anatomy, University of Western Ontario. 'Formerly: Head, Department of Zoology, University of Western Ontario. Present address: Vector Biology and Control, World Health Organization, Geneva, Switzerland. under strong selection with 0.028 ppm fenthion in each generation (LCrO 0.017 ppm fenthion). Thus it had a 7-fold resistance to fenthion; in addition, it had a 10-fold cross-resistance to the S-methyl analogue of fenthion and a 3-fold cross- tolerance to malathion and diazinon.When it was used for crosses with marker strains it was purified by eliminating the yellow larvae. w,y: a susceptible strain of C. p. pipiens homozygous for w (white eye, in linkage- group 1) and y (yellow larva, in linkage- group 2). kps: a susceptible strain of C. p. pipiens homozygous for kps (clubbed palpi, in linkage-group 3): this marker was visible only in males. ru,y: a susceptible strain of C. p. fatigans homozygous for ru (ruby eye) and y, both in linkage-group 2. ru,y, kps: a composite strain obtained by crossing females of the ru,y strain with males of the kps strain. Susceptibility levels were assessed by the standard method for mosquito larvae (WHO Expert Commit- tee on Insecticides, 1963). Test insecticides were not only fenthion but also its S-methyl analogue, 0-methyl S-methyl O-[(4-methylthio)-m-tolyl] phos- phorothioate, hereafter called fenthion-S-methyl. The offspring of crosses were first categorized for visible markers and then for resistance by exposure to a diagnostic concentration in lots of 25 larvae. The characters w and ru were detectable in the 2596 727 C. DORVAL & A. W. A. BROWN heads of the larvae, and it was with kps that it was necessary first to categorize the larvae for resistance and then examine the emerged male survivors for the visible marker. All crosses and backcrosses were made with 150-200 adults of either sex. The possibility of monofactoriality for resistance was assessed in the backcross offspring and the F2 by deriving expected dosage-mortality lines from the combination of the lines for each of the genotypes and then measuring the divergence from them of the observed dosage-mortality relationships by means of chi-square analysis (Georghiou & Barber, 1965). A rough estimate of the number of genes involved (n) was obtained from the difference between parental means (D) and the standard deviation of the F2 (a2) as compared with that of the F1 (al), after the cumulative dosage-mortality lines had been converted to non-cumulative fre- quency distributions; the formula of Castle (1921) derived n by dividing D2 by 8 (a 2-a22). The effect of a given chromosome on resistance, as judged by the marker in that linkage-group, was assessed by factorial analysis as developed for resistance studies by Tsukamoto (1964). Estimations of percentage recombination (crossover) and order of the genes, with corrections for incomplete penetrance, were made by the method of Tsukamoto (1965). The effects of differential viability (,) and partial manifestation (A) of alleles were assessed by the method of Bailey (1961). While the rates of egg hatch in the R and S strains exceeded 90 %, and in the marker strains they lay between 60% and 85%, the hybrid eggs, where the R strain ovum had been fertilized by an S or marker-strain sperm, showed a reduced hatch (30%-70%). The hatch rate recovered to levels of 40%-90% in the backcrosses with the S strain, so that cytoplasmic incompatibility (Laven, 1957) did not impede the crossing experiments. The esterase activities in the R and S strains and the offspring of cross and backcrosses were compared by the agar-gel electrophoresis method of Ogita (1964), as employed by Stone & Brown (1969). On one and the same plate, aliquots of homogenates from R, S, F1 or backcross offspring were applied alongside each other for comparison. Homogenates were made from 0.5 g of larvae in 1 ml of ice-cold phosphate buffer, but with the backcross offspring they were made from single larvae in a minimal amount of buffer. The substrate employed was 2- naphthyl acetate at pH 8.6, and in one case 1-naph- thyl acetate, and the chromogen was Diazoblue B. The density of the single main esterase band was measured with a universal electrophoresis densito- meter (Densicord Model 542) with an integrator connected to the recorder. RESULTS Mode of inheritance offenthion resistance When females of the fenthion-resistant strain were crossed with the susceptible Rangoon strain, the hybrids (Fig. 1) were intermediate in their FIG. I DOSAGE-MORTALITY RELATIONSHIPS TO FENTHION OF THE Sd x RR HYBRID AND OF THE BACKCROSS OFFSPRING FROM ITS RECIPROCAL CROSSES WITH THE S STRAIN 98 90 70 4E W50 Xc 30 10 0.00125 0.0025 0.005 0.01 Fenthion (ppm) . Expected A (Sc x RY) x SS a (So*x RY)4TxSY 0.02 0.04 "o oos9s * (S,3xRy)ixRo o (S&x RM9 x R9 larval response to fenthion, the inheritance being slightly on the recessive side. Similar results were obtained with the reciprocal cross of S females with R males. When fenthion-S-methyl was substituted for fenthion as the test insecticide (Fig. 2), the hybrids from the reciprocal crosses were exactly intermediate in their response. When the S&C x R ? hybrids were backcrossed with the S parent, the dosage-mortality relationships observed for the backcross offspring (Fig. 1) showed 728 INHERITANCE OF RESISTANCE TO FENTHION IN C. P. FATIGANS FIG. 2 DOSAGE-MORTALITY RELATIONSHIPS TO FENTHION-S-METHYL OF THE F. HYBRIDS OF RECIPROCAL CROSSES BETWEEN THE R AND S STRAINS FIG. 3 DOSAGE-MORTALITY RELATIONSHIPS TO FENTHION-S-METHYL OF THE BACKCROSS OFFSPRING FROM THE RECIPROCAL CROSSES OF THE SY x Rd HYBRID WITH THE S STRAIN 0.0625 0.125 0.25 0.5 1.0 Fenthion-s-methyl ( ppm ) 2.0 4.0 --LI oo89s4 a SYix Rd A S x RY an inflexion around the 50% mortality level. Chi- square analysis confirmed that they were not signi- ficantly different from those expected by compu- tation on the basis of a 1 :1 ratio, the X2 values (for 1 d.f.) being only 6.8 and 3.3, and when the SY+RS hybrid was crossed the X2 values between observed and expected for the backcross offspring were only 12.5 and 8.5. Thus the fenthion resistance as assessed by the F, x S backcrosses is essentially monofactorial. With the four F1x R backcrosses, however, the X2 deviations from monofactorial expectations were all in excess of 22.8 and thus significantly different. When, however, fenthion-S-methyl was substituted for fenthion as the test insecticide (Fig. 3), an inflexion was evident at the 50% mortality level. The X2 values were only 4.0 and 4.1 when the S Y x RC hybrids were used, indicating a congruence with monofactoriality; however, when the S,S x R Y hybrids were used, the differences from mono- factoriality were significant. It should be noted that these backcrosses were tested with an old sample of fenthion-S-methyl. e Expected A (Si x R)T x SY & (S Y x Rd)Y x Sc (SY x RQd x RY (Six RF)?x Rl Taken as a whole, the X2 figures for all 16 back- crosses tested indicate that the resistance to fenthion or to its S-methyl analogue is determined by a sin- gle principal gene supported by others. Indications of the approximate number of genes involved were also sought by obtaining the F2 of intercrosses within the F1 hybrids. The F2 obtained from the S?xR,R hybrid (Fig. 4) showed dosage-mortality relationships which, although significantly different from the monofactorial expectation when fenthion was used (X2 23.4), were relatively congruent with it when fenthion-S-methyl was used (X2 8.6). The F2 from the S& xR y hybrid, however, was signi- ficantly different from the monofactorial expectation with both insecticides (X2 16.3 and 21.3, all values for 1 d.f.). The coefficient of variation of the dosage-mortality relationships of the F2 from the S ?x RS hybrids exceeds that of the F1 by 1.6 times (to fenthion) and 2.7 times (to fenthion-S-methyl), indicating 90 r ,1701 e: I . i0 1541 _L I G 104 729 C. DORVAL & A. W. A. BROWN FIG. 4 DOSAGE-MORTALITY RELATIONSHIPS OF THE Ft OFFSPRING FROM INTERCROSSES BETWEEN THE S9 x RS HYBRIDS, TO FENTHION AND TO FENTHION-S-METHYL 98 90 - 70 E D 50 p 30 10 * Expected F2 A Observed F2 in a rough way that more than 1 and less than 3 genes are involved. When the number of factors involved is calculated by Castle's formula, the values of n come out as 4.4 and 1.8, respectively. The concentration of 0.004 ppm was found to be a diagnostic dose separating all the hybrid heterozygotes from the S homozygotes from the kps strain. It was also diagnostic where the w,y strain, the ru,y strain, and the ru,y, kps strain were involved, but with the Rangoon S strain 0.0035 ppm fenthion was the diagnostic concentration. With fenthion-S-methyl, a concentration of 0.35 ppm served as a diagnostic for the w,y and ru,y strains, and 0.40 ppm was diagnostic for the ru,y,kps strain. The success of the diagnostic dose in separating the F1 x S backcross offspring into their 2 types furnishes practical evidence for the single principal gene. This gene allele is therefore given the designa- tion fe. Linkage relationships offenthion-resistance factor The fenthion-resistant strain was crossed with the susceptible marker strain, and the F1 hybrids were backcrossed with the marker strain; the backcross offspring were then categorized for the marker phenotypes by visual examination, and for fenthion resistance by the diagnostic larval dosage. The first exploration of linkage was made by comparing the backcross-offspring phenotype dis- tribution of the survivors (i.e. resistance hetero- zygotes) with that of the backcross offspring not treated with the diagnostic dosage, to ascertain whether the fenthion-resistance character stayed linked with the normal allele of any marker. The crosses were first made with the kps strain marking linkage-group 3, and then with the ru,y, kps strain marking linkage-groups 2 and 3; the possible linkage of fenthion-resistance with sex, associated with linkage-group 1, was also assessed in these crosses. The results (Table 1) show that the factor for fenthion resistance (fe) was not linked with sex or with kps. It was, however, linked with the normal alleles of ru and y in linkage-group 2, the X2 for the comparison showing a highly significant difference between the phenotype distribution of the heterozygotes and that of the untreated backcross- offspring generation as a whole. The linkage is, however, not close, the numbers in each class indicating the crossover value from fe to be 39.7°% for ru and 41.6% for y. The greater X2 value for ru than for y is a second indication thatfe is closer to rni than to y. TABLE 1 COMPARISON OF FREQUENCIES OF EACH PHENOTYPE CLASS BETWEEN THE FENTHION-RESISTANT HETEROZYGOTES (TREATED) AND THE HETEROZYGOTES PLUS SUSCEPTIBLE HOMOZYGOTES (UNTREATED) IN THE OFFSPRING OF BACKCROSSES WITH 2 MARKER STRAINS Marker strain 1 Phenotype Untreated Treated X2 a kps kps: + 630: 548 736: 604 1.47 Male: female 1178:1172 1340:1379 0.76 ru, y, kps kps : + 139 :133 390: 367 0.05 Male: female 272: 279 786: 807 0.00 y: + 265: 286 650: 943 33.88 ru : + 267 : 284 626: 967 53.55 a For I degree of freedom. The linkage of fe with ru was more closely mea- sured by crosses with the ru,y strain, to obtain a total of 10 618 backcross offspring (Table 2). Of these, 4466 were new combinations (ru fe and + +), indicating that the average crossover value was 42.1%. The validity of these results is supported IS?xPd 7 A R ol . .x ol. ,, x 0.00125 0.0025 0.005 0.01 0.02 0.04 0.125 0.25 0.5 1.0 2.0 4.0 -nO ec-96 Fenthion (ppm) Fenthion-s-methyl (ppm) 730 INHERITANCE OF RESISTANCE TO FENTHION IN C. P. FATIGANS TABLE 2 NUMBERS IN EACH PHENOTYPE CLASS IN THE BACKCROSS OFFSPRING a FROM CROSSES WITH THE ru, y STRAIN, WITH THE x2 FOR DEVIATION FROM INDEPENDENT ASSORTMENT |B. B2 B3 B4 + fe 789 484 962 1158 ru fe 626 516 532 788 ++ 536 487 420 561 ru + 651 623 623 862 x2 b 27.6 58.8 144.6 133.4 a B. = (ru,y 3xRY) $xru, y d B2 = (ru, ySx R$) &xru,y V B3= (ru,y xxRS) +xru,yd' B4 = (ru, y Vx R) Sxru,yY b For 3 degrees of freedom; in calculating the expected values, one-quarter of the total was corrected by the observed ratios between ru and + types and between fe and + types, according to the method of Bailey (1961). by the equal numbers of ru and ru+ types, and the almost equal numbers of fe and fe+ types. In this cross, the deviation from independent assort- ment as measured by chi-square is again greater and more consistent for fe-ru than for fe-y. An accurate calculation of the crossover distance fe-y was not possible because y suffers from variable expression and partial manifestation (A). The results of this cross do, however, allow an idea to be obtained of the order of the genes if they are treated by the method of Tsukamoto (1964) for calculating the recombination values after having been corrected for differential viability and for A. When the less likely order fe-y-ru is assumed, the method yields average recombination values of 570% for fe-y, 280% for y-ru, and 45% for fe-ru, thus indicating the order to be fe-ru-y. When the figures are treated for the order fe-ru-y (Table 3), the crossover values were found to be on the average 42% for fe-ru and 27% for ru-y, being greater than these in the male parent and less than these in the female parent. There remained the necessity to assess the possible influence of linkage-group 1 as compared with that of linkage-group 2, and therefore backcrosses were made with the w,y strain. The proportions of phenotypes in the offspring (Table 4) show that fe remained associated with the normal allele of w more often than not. Chi-square analysis as a measure of linkage becomes invalid for these results, TABLE 3 ESTIMATES OF CROSSOVER VALUES, AFTER CORRECTION FOR M AND A, AND CONCLUDING THE ORDER TO BE fe-ru-y Backcross offspring fe-ru ru-y From male hybrid B2 46.64 29.12 B4 39.96 31.49 Average 43.30 30.30 From female hybrid Bi 43.49 16.60 B3 37.76 27.96 Average 40.62 22.28 TABLE 4 NUMBERS IN EACH PHENOTYPE CLASS IN THE BACKCROSS OFFSPRING a FROM CROSSES WITH THE w, y STRAIN Phenotype class B B2 F B3 B4 Total w + + 212 260 551 199 1 222 + fey 556 373 346 291 1 566 w + y 749 755 1 797 594 3 895 + fe + 995 1 078 1 127 514 3 714 Total 2 512 2 466 3 821 1 598 10 397 w fe + 177 280 259 66 782 + + Y 435 208 700 421 1 764 w fe y 246 353 318 262 1 179 + + + 559 462 1 485 492 2 998 Total 1 417 1 303 2 762 1 241 6 723 a See footnote to Table 2. since it even indicated linkage between w and y. This artifact is almost certainly due to the appearance of w types in significantly lower numbers than w+ types, due to differential viability (u). An analysis of variance, which eliminated much of the false linkage, did indicate that the contribution of linkage- group 1 as marked by w was less significant (F = 13.29) than that of linkage-group 2 as marked by y (F = 48.45). 731 C. DORVAL & A. W. A. BROWN TABLE 5 RELATIVE INTENSITIES OF THE ESTERASE BANDS IN THE R STRAIN, THE Ft HYBRID AND THE F, x S BACKCROSS OFFSPRING, AS COMPARED WITH THE S STRAIN Cross [ Homogenate T S F, R Fi x S offspring Sd x R? 0.5-mg. larvae a 1.0 1.26 2.06 1.72 (S&x R ?) c x S single larvae 1.0 1.36 1.75 2.28 2.19 S?xRd 0.5-gm. larvae 1.0 1.23 1.94 1.25 1.91 1.32 1.94 1.34 1.76 (S xRR ) 9 xSd single larvae b 1.0 1.36 1.83 1.08 1.50 0.91 1.46 1.06 1.88 (SV x Rd) d x SV single larvae 1.0 1.32 1.65 1.05 1.23 1.32 a Larval homogenate was centrifuged; remainder of samples in this table were not centrifuged. b The two genotypes In the backcross offspring were first separated by a diagnostic dose before being homogenized. Inheritance of esterase activity Electrophoresis of homogenates, whether centri- fuged or not, confirmed that the R strain was twice as active on 2-naphthyl acetate as the S strain, and the F1 hybrids were one-quarter to one-half more active than the S strain, as judged from plates in which 3 genotypes were tested together in parallel (Table 5). When the F1 x S backcross offspring were obtained and tested on the same plate as their parents, their bands fell into two types corresponding to the S homozygotes and the F1 heterozygotes. When the backcross offspring were divided by the discriminating dose of 0.004 ppm fenthion, the susceptible moiety of larvae being removed as soon as they showed symptoms, the S larvae showed the faint band and the heterozygous resistant larvae showed the band of the intermediate intensity charac- teristic of the F. heterozygotes (Fig. 5). A similar picture was shown when 1 -naphthyl acetate was used as the substrate. DISCUSSION Sufficient evidence to conclude that the fenthion resistance induced in Culex pipiensfatigans is mainly due to a single genetic factor was provided by the F1xS backcrosses when fenthion was used as a diagnostic and by the F1xR backcrosses using fenthion-S-methyl as a diagnostic. This fenthion resistance in this mosquito, as compared with the organophosphorus resistances known in other in- sects, is unusual in not being dominant, the resis- tance ratio shown by the R x S hybrids being 2.1-2.5 times the normal, as compared with a 7-fold resistance in the R strain, which is presumably pure for the resistance factor. The genetic association between fenthion resistance and the main esterase band in the homogenate zymograms is remarkably close, since the intensity in hybrids is 1.25-1.5 times the normal, and in the resistant homozygotes twice the normal. Although not proven, it is likely that 732 INHERITANCE OF RESISTANCE TO FENTHION IN C. P. FATIGANS the fe factor found in linkage-group 2 determines this detoxifying esterase band. The main fenthion-resistance gene was determined to be 42 crossover units from the marker ru in C. p. fatigans, as compared with the crossover dis- tance of 48 units found for genes in the fenthion- resistant (and malathion-resistant) Nanatsumatsu strain of C. p. pallens (Tadano, 1969b). By means of 3-point crosses, Tadano (1969b) obtained evi- dence that malathion resistance in this strain was separated from fenthion resistance by 12-15 units, in the order ma-fe-ru. The crossover distance ma-ru was 48-50 units (Tadano, 1969b), although a distance of 40-43 units had been found before this strain had been purified by selection (Tadano, 1969a). Utilizing the marker y, Tadano (1969a) found in C. p. pallens that the order was ma-ru-y, the distance ma-y being 46-48 units. In these experiments with C. p. fatigans, the order was found to be fe-ru-y, the distance fe-y being 2 units greater than the distancefe-ru. The crossover ru-y was determined to be 17 % from males and 24 % from females, as com- pared with the figures of 22.5% and 23.30% from the two sexes obtained by Tadano & Brown (1967). If the results for fenthion resistance in Culex pipiens s. 1. are compared with those obtained for malathion resistance in Aedes aegypti by Pillai & Brown (1965), it is seen that in both cases the prin- cipal genetic influence came from the linkage-group 2 which contained the DDT-resistance gene. Whereas in Ae. aegypti the additional factor(s) for malathion resistance were in linkage-group 3, in C. p. fatigans this association was found to be with linkage-group 1. That it was associated with w but not with sex may be related to the fact that there is crossing-over between w with sex (Gilchrist & Haldane, 1947). It is possible that the second fenthion-resistance mechanism, namely reduced absorption (Stone & Brown, 1969), derives partly from the genetic influence in linkage-group 1. ACKNOWLEDGEMENTS The authors are most grateful to Dr R. K. Misra for statistical advice, and to Mrs E. Wambera and Mrs G. Riekhoff for valuable technical assistance. They would like to thank Dr C. A. Anderson of the Chemagro Corporation, Kansas City, Kans., USA, for the pure samples of fenthion and its S-methyl analogue. RESUME TRANSMISSION HEREDITAIRE DE LA RESISTANCE AU FENTHION CHEZ CULEX PIPIENS FATIGANS WIED. On a etudie les modalites de la transmission de la resis- tance au fenthion chez une souche de Culex pipiens fati- gans resistante a cet insecticide obtenue par selection a partir d'une souche sensible de Rangoon (Birmanie). On a procede par croisements et retrocroisements avec des souches marqueuses et avec la souche normale. Les croisements r6ciproques entre males et femelles des souches resistantes et sensibles ont donne des resul- tats identiques, demontrant que la resistance au fenthion est un caractere legerement recessif. L'analyse des lignes de regression dose-mortalite relatives aux generations issues de croisements d'hybrides ou de retrocroisements indique que la resistance est sous la dependance d'un gene unique, bien que l'intervention d'autres facteurs genetiques soit manifeste. Le gene principal de la resistance au fenthion (fe) n'est pas lie au sexe; en revanche, il est transmis solidairement avec les alleles normaux des marqueurs ru (ruby eye) et y (yellow larva). I1 est localise dans le groupe de linkage 2, la valeur moyenne de la distance d'enjambement pour ru etant de 40% environ, et les genes etant disposes dans l'ordre fe-ru-y. Une certaine influence du groupe de linkage 1 a ete constatee. Enfin, il existe une correlation etroite entre la resistance au fenthion et l'activite estera- sique sur les acetates de a- et de f-naphtyle, evaluee par electrophorese, d'homogenats de larves sensibles, resis- tantes ou hybrides. 733 734 C. DORVAL & A. W. A. BROWN REFERENCES Bailey, N. J. (1961) Introduction to the mathematical theory ofgenetic linkage. Oxford University Press Castle, W. E. (1921) Science, 54, 223 Georghiou, G. P. & Garber, M. J. (1965) Bull. Wid Hlth Org., 32, 181-196 Gilchrist, B. M. & Haldane, J. B. S. (1947) Hereditas, 33, 175-190 Laven, H. (1937) Z. Vererbungsl., 88, 443-477 Ogita, Z. (1964) Med. J. Osaka Univ., 15, 141-153 Pillai, M. K. K. & Brown, A. W. A. (1965) J. econ. Ent., 58, 255-266 Stone, B. F. & Brown, A. W. A. (1969) Bull. Wid Hith Org., 40, 401-408 Tadano, T. (1969a) Jap. J. exp. Med., 39, 13-16 Tadano, T. (1969b) Jap. J. sanit. Zool., 20, 158-160 Tadano, T. & Brown, A. W. A. (1967) Bull. Wid Hlth Org., 36, 101-111 Tsukamoto, M. (1964) Botyu-Kagaku, 29, 51-59 Tsukamoto, M. (1965) Jap. J. Genet., 40, 159-171 WHO Expert Committee on Insecticides (1963) Wld Hlth Org. techn. Rep. Ser., No. 265, 51-55